CA2769433A1

CA2769433A1 - Biomarkers for early determination of a critical or life threatening response to illness and monitoring response to treatment thereof

Info

Publication number: CA2769433A1
Application number: CA2769433A
Authority: CA
Inventors: Kevin Kain; W. Conrad Liles; Laura Erdman; Andrea Conroy
Original assignee: University Health Network
Current assignee: University Health Network
Priority date: 2012-02-27
Filing date: 2012-02-27
Publication date: 2013-08-27

Abstract

The invention relates to the use of novel biomarkers and biomarker combinations having utility in the early determination of an individual's critical and/or life threatening response to illness and/or in predicting outcome of said illness. The measurement of expression levels of the products of the biomarkers and combinations of biomarkers of the invention have utility in making the determination of an individual's critical and/or life threatening response to illness.
In some embodiments, the biomarker and biomarker combinations are agnostic and are independent of the pre-identification and/or determination of the cause or nature of the illness.
In some embodiments, the biomarkers and biomarker combinations can be utilized to monitor the eflectiveness of treatment interventions for an individual who has a critical illness.

Description

BIOMARKERS FOR EARLY DETERMINATION OF A CRITICAL OR LIFE
THREATENING RESPONSE TO ILLNESS AND MONITORING RESPONSE TO
TREATMENT THEREOF
I. Field of the Invention Encompassed within the scope of the invention is the use of novel biomarkers and biomarker combinations having utility in the early determination of an individual's critical and/or life threatening response to illness and/or in predicting outcome of said illness.
In some embodiments, the biomarker and biomarker combinations are agnostic and are independent of the pre-identification and/or determination of the cause or nature of the illness. In some embodiments, the biomarkers and biomarker combinations can be utilized to monitor the effectiveness of treatment interventions for an individual who has a critical illness.

2. Background of the Invention Diagnosis and Treatment Diagnosis, in the medical context, is the act or process of identifying or determining the nature and/or cause of an illness by identifying the condition(s) (including the diseases and/or injuries) responsible through evaluation of one or more factors which can include patient history, physical examination, review of symptoms and review of data from one or more laboratory tests. While it is not always possible to identify the exact nature or cause of the illness, differential diagnosis may also be utilized in an attempt to eliminate one or more possible causes in order to select the most likely cause.
Once a diagnosis or differential diagnosis has been made, treatment options are considered, and a treatment strategy chosen. In some cases, treatment may begin before diagnosis has been completed (for example, treatment pending receipt of lab results). In other cases, the cause of the illness may remain elusive, but nevertheless treatment is selected on the basis of the symptoms which the individual presents. When the diagnosis, differential diagnosis, or symptoms are indicative of a condition which has the potential to be critical and/or life threatening, the management strategy may include additional considerations to ensure the best possible clinical outcome including rapid triage, referral, admission to hospital, enhanced monitoring, admission to an intensive care unit, and the like.
The Agnostic Approach to Diagnosis and Treatment The traditional model of selecting a treatment strategy based solely on the pre-determined origin or cause of the illness has some significant drawbacks.
While identifying the cause helps to ensure that the selected course of treatment is disease, injury, or at least symptom specific, it often fails to recognize the importance that the individual's unique response to their condition plays in defining the course and severity of the illness.
The "agnostic" approach to treatment challenges the traditional paradigm of selecting a treatment strategy based on the origin or cause of illness. The agnostic approach is chosen not necessarily because the cause or origin is unknowable (as in the religious context), or because diagnosis cannot be of assistance, but because knowing as early as possible whether an individual will respond critically and/or in a life threatening manner to illness can provide a more effective and rapid method to triage and select treatment tailored to the individual.
Individual's Response to Illness It is well recognized that not all individuals respond to an illness in the same manner. Many develop only mild and self-limited disease, while a small proportion may progress to a critical and/or life threatening stage. At presentation to medical care, it can be difficult to determine who will do well without intervention, or with only minimal intervention, and who needs admission and specialized management in order to improve clinical outcome. For example, in the case the H IN I influenza pandemic, it was estimated that approximately 61 million individuals in the United States were infected with H1N1 (during the period from April 2009 to April 2010), but only a small percentage of those cases resulted in death. Of the 61 million individuals infected, approximately 274,000 individuals were admitted for hospitalization (0.449%), and 12,470 thousand deaths occurred (0.012%) (Emerging Infection Programs Data released May 14, 2010 from the Centre for Disease Control',). Clearly some individual's were more able to fight the H IN1 infection than others.
I Deaths rounded to the nearest ten. Hospitalizations have been rounded to the nearest thousand and cases have been rounded to the nearest million.

Despite this diversity of response, it has been difficult, even with retrospective analysis, to determine what specific factors and characteristics contributed to the differential outcome in these individuals. For example, a retrospective study was performed on worldwide data available prior to July 16, 2009 on the 684 deaths reported as of that date (Valliant, L. et al., Eurosurveillance, Vol. 14, Issue 33, p.1-6 (2009)) and the age of the patients were reviewed, by country. In that study it was found that while overall most deaths (51%) occurred in the age group of 20-49, the impact of age, and the age group most impacted varied in different countries, making it difficult to draw predictive conclusions.
Another example of an illness which has life threatening potential is sepsis (septicemia). Sepsis is a systemic inflammatory response to a presumed infection, and may result from numerous diverse diseases or etiologies. In some cases severe sepsis may develop wherein the syndrome is also associated with organ dysfunction, hypoperfusion, or hypotension.
Because only a small fraction of individuals with an illness proceed to have a critical and/or life threatening response, an ability to differentiate those individuals who require urgent triage and intensive treatment from those individuals who do not, would be of significant advantage.
Current attempts to selectively treat individuals who are most vulnerable for a life threatening response to an illness occurs by first diagnosis said illness, and then either pre-classifying individuals based on known risk factors (e.g. age, existing co-morbidities and the like) and/or by monitoring individuals for early indications that suggest the illness is proceeding in a life threatening manner. For example, a prospective cohort study conducted in 2 phases at 2 general hospitals in Brazil found that by increased monitoring of in-hospital patents using currently existing measurable indicators for detection specific to sepsis, and providing treatment accordingly, the mortality rate for patients was reduced from 61.7% to 38.2% (Wesphal, G.A., et al. "Reduced mortality after the implementation of a protocol for the early detection of severe sepsis" Journal of Critical Care (2011) 26 p.76-81).
Nevertheless, reliance on risk factors remains vastly inadequate as a means of selecting individuals who are likely to have a life threatening response (see Vaillant, L. et al.
supra), and existing measurable indicators that an individual is having a life threatening response often requires extensive and costly monitoring of patients and can take too long to be of clinical use in managing the patient. Furthermore, relying on diagnosis prior to monitoring or providing treatment can increase costs and cause unnecessary delay. This is problematic, particularly in cases where resources are limited, such as in developing

3 countries, but applies equally to developed countries given the costs associated with critical care.
For example, in the case of H1N1 treatment, Durben et. al. modeled the costs from a societal perspective for the treatment of the Ontario population (assuming no preventative vaccination) and determined a total cost of $1.10 billion dollars with approximately 87 million dollars being allocated to various aspects of hospital care (Durben et al. (2011) "A cost effectiveness analysis of the H1N1 vaccine strategy for Ontario, Canada" Journal of Infectious Diseases and Immunity Vol. 3(3) p. 40-49). The early and accurate identification and stratification of those individuals more likely to have a poor response to the infection could have focused resources on those most likely to benefit from them and away from the majority of infected individuals who recovered well without specific medical intervention. This strategy would presumably have decreased these projected costs quite significantly.
Thus, what is needed in the art is one or more biomarkers which provide greater certainty than current models of an individual's increased risk of progressing to a critical and/or life threatening response to illness so as to select and/or modify a treatment protocol for said individual. Preferably these biomarkers would recognize the increased risk as early as possible so as to allow the greatest potential for treatment intervention. It would also be particularly helpful if the biomarkers were agnostic and had utility irrespective of the illness, so it would be unnecessary to first diagnose the illness. Also, the ability to use one or more biomarkers to monitor the impact of the treatment protocol on the progress of a life threatening response would permit modification of the treatment protocol as necessary would also be of significant benefit.
3. Summary In one aspect, what is disclosed are biomarkers and biomarker combinations which provide an indication of an individual's response to illness, the severity of that response, and whether they already have, or are progressing to, a critical and/or life threatening form of illness. In another aspect the biomarker and biomarker combinations are capable of providing an early indication of the severity of an individual's response to illness which is not predicated upon first determining the cause or source of the illness. In yet another aspect, what is disclosed

4 are biomarkers and biomarker combinations which provide an early indication of the impact of the treatment protocol on the individual's risk or progress of their life threatening response.
In another aspect is a composition comprising a collection of two or more antibodies and a suitable buffer, the composition capable of selectively binding to at least two protein biomarkers from a sample isolated from a test individual, where the protein biomarkers are those in Table 1. In another aspect the composition is a composition comprising three or more antibodies and the composition is capable of selectively binding to at least three protein biomarkers from a sample isolated from the test individual, where the protein biomarkers are those in Table 1. In another aspect the composition comprises a collection of two or more antibodies and a suitable buffer, the composition is capable of selectively binding to at least two protein biomarkers from a sample isolated from a test individual, and the protein biomarkers are C5a, VEGF, sFlt-1, CHI3L I, CRP, Ang-like3, FactorD, or ILI 8bp). In another aspect the composition comprises a collection of three or more antibodies and a suitable buffer, the composition is capable of selectively binding to at least three protein biomarkers from a sample isolated from a test individual, and the protein biomarkers are C5a, VEC;F, sFlt-1, CH131_,1, CRP, Ang-like3, FactorD, or IL18bp). In yet another aspect, the sample is a whole blood sample, a serum sample or a plasma sample.
In some embodiments, the compositions are used to (i)detect and quantify a level of the two or more protein biomarkers in the sample, (ii)compare the quantified level to control levels of the protein biomarkers in a control population, (iii) determine the presence of differential levels for the two or more biomarkers so as to make a determination that the individual is at a significantly increased risk of having a critical and/or life threatening response to illness as compared with the control population. In some embodiments, the detecting and quantifying utilizes one or more devices to transform the sample into data indicative of the levels of each of the two or more protein biomarkers. In some embodiments, the device is an enzyme linked immunoassay which is utilized to transform the sample into data. In some embodiments, the test individual is subjected to a treatment protocol on the basis of the determination in step (iii).
in some embodiments, the control population is an unbiased population of individuals having the same illness as the test individual. In some embodiments, the control population is a population of individuals having the same illness as the test individual, and not developing a critical and/or life threatening response to the illness. In some embodiments, the control population is a population of individuals who are normal. In some

5 embodiments, the control population is a population of individuals wherein the majority of members of the control population do not have the same illness as the test individual.
In some embodiments, there is a method of determining the likelihood that a test individual has or will develop a critical and/or life threatening response to illness, where the method includes (i) detecting and quantifying a level of each of two or more protein biomarkers in a sample, where the protein biomarkers are those in Table 1 (ii) comparing the quantified levels of said protein biomarkers to control levels of the protein biomarkers from a control population (iii) determine the presence of differential levels for the two or more biomarkers based on the comparison in step (ii) so as to make a determination that the individual is an increased risk of having a critical and/or life threatening response to illness when compared with the control population.
In some embodiments, the determination is made that the individual is at a significantly increased risk. In some embodiments, the detecting and quantifying of step (i) utilizes one or more devices to transform the sample into data indicative of the levels of each of the two or more protein biomarkers. In some embodiments, the one or more devices is an enzyme linked immunoassay. In some embodiments, the individual is subjected to a treatment protocol on the basis of the determination made. In some embodiments, the control population is an unbiased population of individuals having the same illness as the test individual. In some embodiments, the control population is a population of individuals having the same illness as the test individual, and not developing a critical and/or life threatening response to the illness. In some embodiments, the control population is a population of individuals who are normal. In some embodiments, the control population is a population of individuals wherein the majority of members of the control population do not have the same illness as the test individual.
In some embodiments, there is a method of determining the likelihood that a test individual will develop a critical and/or life threatening response to illness, where the method includes (i) detecting and quantifying a level of each of two or more protein biomarkers in a sample, where the protein biomarkers are those in Table 1 (ii) using the quantified levels of each of the protein biomarkers from the sample in a classifier where the classifier was generated using two populations, a first population who developed a critical and/or life threatening response to illness and a second control population, (iii) making a determination as to whether the quantified levels are indicative of the individual being more similar to the first population or the second control population so as to determine whether the

6 individual is at an increased risk of developing a critical and/or life threatening response to ii iness.
In some embodiments, the determination is made that the individual is at a significantly increased risk. In some embodiments, the detecting and quantifying of step (i) utilizes one or more devices to transform the sample into data indicative of the levels of each of the two or more protein biomarkers. In some embodiments, the one or more devices is an enzyme linked immunoassay. In some embodiments, the individual is subjected to a treatment protocol on the basis of the determination made. In some embodiments, the second control population is an unbiased population of individuals having the same illness as the test individual. In some embodiments, the second control population is a population of individuals having the same illness as the test individual, and not developing a critical and/or life threatening response to the illness. In some embodiments, the second control population is a population of individuals who are normal. In some embodiments, the second control population is a population of individuals wherein the majority of members of the control population do not have the same illness as the test individual.
In some embodiments, the test individual has not been diagnosed or differentially diagnosed with an illness which has the potential to become critical and/or life threatening prior to use of compositions or methods as disclosed.
4. Brief Description of the Drawings The objects and features ofthe invention can be better understood with reference to the following detailed description and drawings.
Figure IA and IB in one embodiment, compares protein biomarker levels isolated from plasma in children who have been diagnosed as having malaria (including individuals who can be subclassified as having either cerebral malaria (CM) or severe malarial anemia (SMA)) and who survived the malaria, as compared with the protein biomarker levels isolated from plasma in children who died from the malaria and demonstrate a statistically significant difference as between the two phenotypic groups. Figure IA shows the results from biomarker Ang-2, sICAM-1, sFlt-1, CF1131,1, IP-10, sTie-2, and PCT.
Figure IB
shows the results from biomarker sTREM-I. * indicates a statistical difference in the protein levels with a p value of < 0.05. ** indicates p values of < 0.01.
Figure 2A, in one embodiment, demonstrates the receiver operating characteristic (ROC) curves generated using the selected biornarkers sICAM-I, sFlt-1, Ang-2, PCT,

7 sTREM-1, and CI-1131-1 to differentiate between fatal and non-fatal malaria.
Dashed reference lines represent the ROC curve for a test with no discriminatory ability. Area under the ROC curve is noted in each graph with the 95% confidence interval shown below in parentheses. P values are indicated * p<0.05, ** p<0.01.
Figure 2B, in one embodiment, demonstrates the receiver operating characteristic (ROC) curve for parasetimia diagnosis alone. Dashed reference lines represent the ROC curve for a test with no discriminatory ability. Area under the ROC
curve is noted in each graph with the 95% confidence interval shown below in parentheses. P
values are indicated * p<0.05.
Figure 3, in one embodiments, demonstrates a classification tree analysis used to predict outcome of severe malaria infection with host biomarkers where six biotnarkers were entered into the CRT, and the resulting CRT using IP-10, Ang-2, and sICAM-1 resulted as shown with cut-off points as determined. Prior probabilities of survival and death were specified (94.3% and 5.7% respectively). The cut-points selected by the analysis are indicated between parent and child nodes. Below each terminal node (ie no further branching), the predicted categorization of all patients in that node is indicated. The model yields 100% sensitivy and 92.5% specificity for predicting mortality (cross validated misclassification rate 15.4% with standard error 4.9%).
Figure 4A, in one embodiment, demonstrates the absolute and median concentrations of angiopoietin-1 (Ang-1) and angiopoietin-2 (Ang-2), as well as the ratio between the two (Ang-2:Ang-1 expressed as log base 10) in acute and convalescent plasma from patients with or without STSS. * P < 0.05; ** P <0.01.
Figure 4B, in one embodiment, demonstrates the receiver operating characteristic curves for each of Aug-1, Ang-2 and the ratio between the two, comparing patients with STSS in the acute phase of illness to those without STSS, also in the acute phase of illness.
Figure 5, in one embodiment, shows Angiopoietin-1 and -2 (Ang-1 and Ang-2) concentrations, and the ratio between the two (Ang-2:Ang-I), in matched acute and convalescent plasma samples from patients with invasive Group A streptococcal infection and STSS.
Figure 6A, in one embodiment, is a histogram showing the relationship between mortality (%) and measured Aug-1 levels on admission.

8 Figure 6B in one embodiment, shows a receiver operating characteristic (ROC) curve illustrating added sensitivity and specificity in predicting 28-day mortality when comparing plasma Ang-1 levels, MOD score or age with the combination of the three variables.
Figure 7A, in one embodiment, shows the comparison of Ang-2 levels with MOD
score as predictors of mortality in patients with severe sepsis.
Figure 7B, in one embodiment, shows the comparison of Ang-2 levels taken one day prior to assessing the MOD score in patients with severe sepsis.
Figure 8A, in one embodiment, shows the levels of Angiopoietin-1 (Ang-1), Angiopoietin-2 (Ang-2) and the Ang-2:Ang-1 ratio in children with uncomplicated E. coil 0157:117 infection (infected), children prior to the diagnosis of HUS (pre-HUS), and children demonstrating HUS at the time of diagnosis (HUS). *p<0.05, "p<0.01 unfilled circles indicate outliers (1.5x interquartile range (IQR], filled circles indicate extreme outliers (3x1QR).
Figure 88, in one embodiment, shows Receiver Operating Characteristic (ROC) curves for Ang-1, Ang-2 and Ang-l:Ang-2 ratio as comparing children with uncomplicated infection and those with the pre-HUS phase of illness, with the null hypothesis being that the area under the curve is 0.5 p---0.01 for Ang-1 5. Detailed Description 5.1 Definitions The following definitions are provided for specific terms which are used in the following written description.
As used herein, the "amino terminal region of a polypeptide" refers to the polypeptide sequence of a protein biomarker. As used herein, the "amino terminal region"
refers to a consecutive, or nearly consecutive stretch of amino acids located near the amino terminus of a polypeptide and is not shorter than 3 amino acids in length and not longer than 350 amino acids in length. Other possible lengths of the "amino terminal"
region of a polypeptide include but are not limited to 5, 10, 20, 25, 50, 100 and 200 amino acids.
The term "antibody" encompasses monoclonal and polyclonal antibodies and also encompasses antigen-binding fragments of an antibody. The term "antigen-binding fragment" of an antibody (or simply "antibody portion," or "antibody fragment"), as used herein, refers to one or more fragments of a full-length antibody that retain the ability to specifically bind to a polypeptide encoded by one of the genes of a biomarker of the

9 invention.. Examples of binding fragments encompassed within the term "antigcn-binding fragment" of an antibody include (i) a Fab fragment, a monovalent fragment consisting of the VL, VH, CL and CH I domains; (ii) a F(ab')2 fragment, a bivalent fragment comprising two Fab fragments linked by a disulfide bridge at the hinge region; (iii) a Fd fragment consisting of the VH and CHI domains; (iv) a Fv fragment consisting of the VL and VH
domains of a single arm of an antibody, (v) a clAb fragment (Ward et al., (1989) Nature 341:544-546), which consists of a VH domain; and (vi) an isolated complementarity determining region (CDR). Furthermore, although the two domains of the PI/ fragment, VL and VH, are coded for by separate genes, they can be joined, using recombinant methods, by a synthetic linker that enables them to be made as a single protein chain in which the VL and VH
regions pair to form monovalent molecules (known as single chain Fv (scFv); see e.g., Bird et of. (1988) Science 242:423-426; and Huston et of. (1988) Proc. Natl. Acad. Sci. USA
85:5879-5883).
Such single chain antibodies are also intended to be encompassed within the term "antigen-binding fragment" of an antibody. These antibody fragments are obtained using conventional techniques known to those with skill in the art, and the fragments are screened for utility in the same manner as are intact antibodies. The antibody can be monospecific, e.g., a monoclonal antibody, or antigen-binding fragment thereof. The term "monospecific antibody" refers to an antibody that displays a single binding specificity and affinity for a particular target, e.g., epitope. This term includes a "monoclonal antibody"
or "monoclonal antibody composition," which as used herein refer to a preparation of antibodies or fragments thereof of single molecular composition.
As used herein an "array" contemplates a set of protein biomarkers, or antibodies complementary to protein biomarkers, or combinations thereof immobilized to a support. An array can also include fragments of protein biomarkers or fragments of antibodies immobilized to a support wherein the fragment still allows the selective binding of the protein or antibody fragment to its complementary binding partner.
As used herein, the "carboxy terminal region of a polypeptide" refers to the polypeptide sequences of a protein biomarker. As used herein, the "carboxy terminal region"
refers to a consecutive, or nearly consecutive stretch of amino acids located near the carboxy terminus of a polypeptide and is not shorter than 3 amino acids in length and not longer than 350 amino acids in length. Other possible lengths of the "amino terminal"
region of a polypeptide include but are not limited to 5, 10, 20, 25, 50, 100 and 200 amino acids. The "earboxy terminal" region does not normally include the polyA tail, if one is present in the protein biomarker.
As used herein, the term "classifier" includes a mathematical model generated on its ability to differentiate between at least two different traits with respect to an individual's response to illness. Classifiers can include logistic regression, classification tree analysis, or other known mathematical models, and are generated using at least two populations wherein the phenotype of the populations is known. In some embodiments, a first population has been confirmed as demonstrating a critical and/or life threatening response to illness, and the second population is a control population as defined herein. The classifier, so generated, can be used with data from a test individual to generate a numerical output which is indicative of whether the individual is at risk of developing a critical and/or life threatening response to illness, (or is already developing a critical and/or life threatening response to illness), or not.
As used herein the term "complementary binding partner" includes a compound which selectively binds to a protein biomarker and includes nucleic acid aptamers, peptide aptamers, a peptibody, a mimetic, an inhibitor, and any compound that binds to the protein biomarker in vivo, an antibody including a monoclonal and/or polycional antibody.
As used herein the term "control population" is considered in reference to the test individual since the levels of the biomarker and biomarker combinations in the test individual must be compared to levels in the control population to determine the likelihood of the test individual having a critical and/or life threatening response, and/or to predict the outcome of the response. Control populations can either be negative control populations or positive control populations. In some embodiments, the control population is a negative control population, the test individual has been diagnosed with an illness, and the control population is a population of individuals who have had the illness of the test individual and have not developed a critical or life threatening response. In some embodiments, the test individual has been diagnosed with an illness and the control population is a population of normal individuals. In some embodiments, the test individual has been diagnosed with an illness and the control population is an unbiased population of individuals with said illness.
In some embodiments, the control population is a positive control population, the test individual has been diagnosed with an illness, and the control population is a population of individuals who have had the illness and have developed a critical or life threatening response.
Ii In some embodiments, the utility of the biornarkers and biomarker combinations is independent of the cause or source of the illness of the test individual.
Control populations can still either be negative control populations or positive control populations. In some embodiments the test individual has not been diagnosed with an illness prior to testing the biomarker and/or biomarker combinations and the control population is a negative control population of individuals who have had an illness and have not developed a critical and/or life threatening response. In these embodiments, the illness does not have to be the same as the illness of the test individual. In some embodiments, none of the members of the control population have had the same illness as the test individual, In yet other embodiments, the majority of the members of the control population have not had the same illness as the test individual. In yet other embodiments 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90% or more of the control population does not have the same illness as the test individual. In yet other embodiments 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90% or more of the control population has the same illness as the test individual. In some embodiments.
IS the test individual has not been diagnosed with an illness prior to testing the biomarker and/or biornarker combinations and the control population is a population of normal individuals. In some embodiments, the test individual has not been diagnosed with an illness prior to testing, and the control population is a positive control population of individuals who have had an illness and have developed a critical or life threatening response to said illness. In some embodiments, none of the members of the control population have had the same illness as the test individual. In yet other embodiments, the majority of the members of the control population have not had the same illness as the test individual. In yet other embodiments 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90% or more of the control population does not have the same illness as the test individual. In yet other embodiments 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90% or more of the control population has the same illness as the test individual.
As used herein -diagnosis" refers to the act or process of identifying or determining the nature and/or cause of an illness by identifying the condition(s) (including the diseases and/or injuries) responsible through evaluation of one or more factors which can include patient history, physical examination, review of symptoms and review of data from one or more laboratory tests.
As used herein "diagnosed with an illness" refers to having confirming the nature and/ or cause of the illness by identifying the agent, disease, or injury responsible for one or more of the symptoms exhibited by said individual.

As used herein "differentially diagnosed with an illness" refers to having narrowed down the nature and/or cause of the illness sufficiently to ensure that the patient will receive the same treatment that the patient would have received if the nature and/or cause of the illness was known.
As used herein, "illness" refers to a condition which has as one possible outcome a critical and/or life threatening outcome including death. In some embodiments, illness encompasses disorders of endothelial cell function. In some embodiments, illness is one which results from an infection such as a parasitic infection, a viral infection, a bacterial infection, and/or results from bioactive molecules including microbial toxins.
In some embodiments illness includes conditions wherein one of the causes of the condition is a significant burn or physical trauma. In other embodiments illness includes exposure to a biothreat agent such as anthrax. In other embodiments illness includes exposure to agents which can cause acute lung injury, such as smoke,. In other embodiments an illness can include disease caused by weaponized microbes and/or biothreat agents, in some embodiments which cannot be diagnosed using traditional diagnosis techniques.
For example, the virulence factor or toxin of the microbe and/or biothreat agent has been modified and inserted into a harmless carrier bacteria, virus or other carrier agent (Trojan horse effect). Examples of illnesses include but arc not restricted to pneumonias and lower respiratory track infections, influenza, E. coli infections and its complications such as hemolytic uremic syndrome, bacteremias, rickettsial infections, salmonellosis, streptococcal infections, staphylococcus infections, malaria, sepsis, Dengue fever, west nile virus, toxic shock syndrome, leptospirosis, agents causing viral hemorrhagic fever (e,g, Ebola, Marburg), and microbes or biothreat agents, including those that have been altered to obscure traditional diagnosis.
"Differential levels" refers to protein biomarker levels which demonstrate a statistically significant difference in the level when compared with the levels of the protein biomarker in a control population, wherein the difference is at least 10% or more, for example, 20%, 30%, 40%, or 50%, 60%, 70%, 80%, 90% or more or 1.5 fold, 2 fold, 2.5 fold, 3.0 fold, 3.5 fold, or more in protein levels relative to the levels in a control population.
Differentially increased levels" refers to protein biomarker levels which demonstrate a statistically significant increased level when compared with the levels of the protein biomarker in a control population, wherein the increase in levels is at least 10% or more, for example, 20%, 30%, 40%, or 50%, 60%, 70%, 80%, 90% or more or 1.5 fold, 2 fold, 2.5 fold, 3.0 fold, 3.5 fold, or more increase in protein levels relative to the levels in a control population.
"Differentially decreased levels" refers to protein biomarker levels which demonstrate a statistically significant decreased level when compared with the levels of the protein biomarker in a control population, wherein the decrease in levels is at least 10% or more, for example, 20%, 30%, 40%, or 50%, 60%, 70%, 80%, 90% or more or 1.5 fold, 2 fold, 2.5 fold, 3.0 fold, 3.5 fold, or more decrease in protein levels relative to the levels in a control population.
As used herein "an individual's response to illness" indicates an individual's ability to garner resources to control and/or battle the illness and determines the course of the illness within the individual. The individual's response to illness can be influenced by their innate and acquired immune response, genetic background, medical history, health status, age, sex, and pre-existing or co-existing illnesses and/or treatments. In addition, the course of the illness is also affected by the treatment protocol applied for the illness itself. Irrespective of the specific factors which influence the individual's response to illness, the response impacts the course of the illness in that individual.
As used herein "a critical and/or life threatening response to illness" is indicative of an individual's response to the illness such that the individual is at an increased risk of death as compared with the risk of death in an unbiased population of individuals who suffer the illness. In some embodiments the increased risk of death is a "significantly increased risk" which means that the increase in risk as compared to an unbiased population of individuals having the illness is greater than 50%, 60%, 70%, 80%, 85%, 90%, 95% or more.
As used herein, the "internal region of a polypeptide" refers to the polypeptide sequences of a protein biomarker. As used herein, the "internal region" refers to a consecutive, or nearly consecutive stretch of amino acids located within the internal region of a polypeptide and is not shorter than 3 amino acids in length and not longer than 350 amino acids in length. Other possible lengths of the "internal" region of a polypeptide include but are not limited to 5, 10, 20, 25, 50, 100 and 200 amino acids.
As used herein, "normal" refers to an individual, a group of individuals, or a population of individuals who have not shown any symptoms of illness as defined herein and/or do not have an illness.
As used herein, -patient" or "individual" refers to a human.

As used herein, "protein biomarker" refers to the form of the protein, including fragments, which are expressed and potentially processed and exist in sufficient quantity and for sufficient time so as to be capable of being measured in humans using a compound which selectively binds to the protein. Biomarkers may be capable of being used individually, or in combination with other biornarkers, additively or synergistically to provide information as to an individual's response to illness. As used herein "protein biomarker fragments" may include the "amino terminal region of a polypeptide", the "carboxy terminal"
region of a polypeptide" or the "internal polypeptide region of a polypeptide"
As used herein, the terms "purified" in the context of a protein biomarker and/or an complementary binding partner (e.g., a peptide, polypeptide, protein or antibody) refers to a compound which is substantially free of cellular material and in some embodiments, substantially free of heterologous agents (i.e., contaminating proteins) from the cells or tissue source from which it is derived, or substantially free of chemical precursors or other chemicals when chemically synthesized. The language "substantially free of cellular material" includes preparations of a proteins in which the proteins are separated from cellular components of the cells from which it is isolated or recombinantly produced.
Thus, a compound that is substantially free of cellular material includes preparations of a compounds having less than about 30%, 20%, 10%, or 5% (by dry weight) of heterologous proteins (e.g., protein, polypeptide, peptide, or antibody; also referred to as a "contaminating protein").
When the compound is recombinantly produced, it is also preferably substantially free of culture medium, i.e., culture medium represents less than about 20%, 10%, or 5% of the volume of the protein preparation. When the compound is produced by chemical synthesis, it is preferably substantially free of chemical precursors or other chemicals, i.e., it is separated from chemical precursors or other chemicals which are involved in the synthesis of the compound. Accordingly, such preparations of a compound have less than about 30%, 20%,

10%, 5% (by dry weight) of chemical precursors or compounds other than the compound of interest.
As used herein, the term "selectively binds" refers to the specific interaction between a protein biomarker and complementary binding partner which is able to interact with the protein biomarker in specific manner, and preferentially to other proteins. Selective binding of a protein biomarker and a complementary binding partner and includes the specific interaction of an antibody with a protein biomarker, including the binding of a monoclonal antibody and/or a polyclonal antibody to a protein biomarker preferentially in comparison to non-specific binding. Selective binding can also include binding between the protein bioinarker and a nucleic acid or peptide aptamer, a peptibody, or the like. For example, a region, portion or structure of a first protein molecule recognizes and binds to a region, portion or structure on a second protein molecule preferentially to the binding of a non-specific third protein. "Selective binding", "Selective binding", as the term is used herein, means that a molecule binds its specific binding partner with at least 2-fold greater affinity, and preferably at least 10-fold, 20-fold, 50-fold, 100-fold or higher affinity than it binds a non-specific molecule.
As used herein, the term "a therapeutic protocol", refers to a treatment and/or monitoring strategy which an individual is subjected to, and can be as a result of traditional diagnosis, differential diagnosis, identification of symptoms and/or as a result of use of the protein bioinarkers of the invention and can include the application of one or more drug therapies or strategies, medical monitoring which can include increased nursing care, admission to hospital or clinic, admission to an intensive care unit, and or combinations thereof By "an unbiased population" as used herein is meant a population of individuals who have a specific illness, but have not been pre-selected on the basis of one or more known risk factors for response to the specific illness (for example, age, sex, existing co-morbidities and the like).
As used herein, "a plurality of" or "a set or' refers to more than two, for example, 3 or more, 4 or more, 5 or more, 6 or more, 7 or more, 8 or more, 9 or more 10 or more etc.
As used herein, the terms "treat", "treatment" and "treating" refer to the reduction or amelioration of the progression, severity and/or duration of episodes and/or Symptoms of illness.
5.2 Detailed Summary We have reviewed various illnesses, each of distinctly different etiologies, which nevertheless have in common the potential to progress to a stage which is critical and/or life threatening. Another commonality amongst these illnesses is the fact that not all individuals, despite being properly diagnosed, progress to the critical and/or life threatening form of the illness. Although it has been known that the individual response to illness plays a significant role in disease progression, it has been difficult to accurately predict which individuals will demonstrate a critical and/or life threatening response, even once the illness has been diagnosed. We have surprisingly identified certain proteins biomarkers, many of which are involved in endothelial activation and/or inflammation, that are found circulating in the blood of individuals who progress to the critical and/or life threatening stage of illness at different levels than the biomarkers are found in individuals who will not demonstrate a critical and/or life threatening response to illness. The biomarkers are often found at different levels even in the very early stages of illness, and often before other known indicators of disease severity can be measured. More surprisingly, we have found that these biomarkers have utility across a diverse group of illnesses suggesting that these biomarkers have utility even if the individual has not yet been diagnosed or differentially diagnosed with a specific illness, making the application of these biomarker particularly useful in situations where:
diagnosis is not possible (such as in cases of weaponized microbes or biothreat agents which have been designed to prevent identification), diagnosis may be too costly (such as in developing worlds), diagnosis can delay appropriate treatment, or diagnosis results in overabundance of treatment. As such, we have identified proteins that represent early indicators that an individual is unable to respond effectively to illness and will progress to a critical and/or life threatening stage of illness. Because these proteins are differentially found across such diverse diseases, they have the ability to be used apriori to diagnosis allowing more timely and cost effective interventions than would otherwise be available.
The practice of the present invention employs, in-part conventional techniques of protein chemistry and molecular biology which are within the skill of the art. Such techniques are explained fully in the literature. See, e.g., Sambrook, Fritsch & Man iatis, 1989, Molecular Cloning: A Laboratory Manual, Second Edition ; Oligonueleotide Synthesis (Mi. Gait. ed., 1984); Nucleic Acid Hybridization (B.D. Flames & S.J. Higgins, eds., 1984);
A Practical Guide to Molecular Cloning (B. Perbal, 1984); and a series, Methods in Enzymology (Academic Press, Inc.); Short Protocols In Moleeular Biology, (Ausubel et al., ed., 1995). All patents, patent applications, and publications mentioned herein, both supra and infra, are hereby incorporated by reference in their entireties.
5.3 Control and Test Samples In some embodiments, all that is required is a drop of blood. This drop of blood can be obtained, for example. from a simple pinprick. In some embodiments, any amount of blood is collected that is sufficient to detect the expression of one, two, three, four, five, six, seven or more of the genes in Table I. In some embodiments, the amount of blood that is collected is I ul or less, 0.5 ul or less, 0.1 ul or less, or 0.01 ul or less. In some embodiments more blood is available and in some embodiments, more blood can be used to effect the methods of the present invention. As such, in various specific embodiments, 0.001 ml, 0.005 ml, 0.01 ml, 0.05 ml, 0.1 ml, 0.15 ml, 0.2 ml, 0.25 ml, 0.5 ml, 0.75 ml, 1 ml, 1.5 ml, 2 ml, 3 ml, 4 ml, 5 ml, 10 ml, 15 ml or more of blood is collected from a subject. In another embodiment, 0.001 ml to 15m1, 0.01 ml to 10 ml, 0.1 ml to 10 ml, 0.1 ml to 5 ml, Ito 5 ml of blood is collected from a subject.
In some embodiments, whole blood is utilized. In some embodiments of the present invention, whole blood collected from a subject is fractionated (i.e., separated into components) and only a particular fraction is utilized. In some embodiments only blood serum is used, wherein the serum is separated from the remaining blood sample by isolating the liquid fraction of blood which has been allowed to clot. In some embodiments plasma samples are used, wherein the blood has been pre-treated with an anticoagulant, such as EDTA, sodium citrate (including buffered or non-buffered), heparin, or the like and the supernatant collected and utilized. In some embodiments, the blood is subjected to Flea-Hypaque (Pharmacia) gradient centrifugation and the peripheral blood mononuclear cells (PBMC's) are used. Other fractions and/or fractionating techniques known in the art may also be used, for example, blood cells can be sorted using a using a fluorescence activated cell sorter (FACS) e.g. Kamarch, 1987, Methods Enzymol 151:150-165).
5.4 Biomarker and Biomarker Combinations Table I provides a list of proteins which are useful as biomarkers either individually or in combination.
The biomarkers may be used to determine an individual's status with respect to their developing a critical and/or life threatening response to illness. In some cases the biomarkers are individually useful in helping to assess the likelihood of an individual having a critical and/or life threatening response to illness. In some cases the biornarkers are useful in helping to assess whether an individual is at a significantly increased risk of a critical and/or life threatening response. In yet other instances the biomarkers are useful in helping to assess whether an individual is not at a significantly increased risk of having a critical and/or life threatening response. In yet other instances, the biomarkers are useful in assessing the impact of a treatment protocol on an individual who has a significantly increased risk of a critical and/or life threatening response. In some cases, the biomarkers are useful in =

determining the likelihood of an individual demonstrating an improvement in their critical and/or life threatening response.

'-.`-411,fiqq04%-4:0150.40,4 Complement fragment C5a C5a Angiopoietin-1 Ang- I
Angiopoietin-2 Ang-2 kDa interferon gamma-induced protein 1P-10 Soluble intercellular adhesion molecule- I sICAM-1 Vascular endothelial growth factor A VECTF
soluble Fins-like tyrosine kinase receptor-I (also known as soluble VEGFR1 -Vascular Endothelial Growth Factor geceptor 1) sFit-1 Chitinase-3-like protein 1 CH13L1 Soluble triggering receptor expressed on myeloid cells-1 STREM-1 C-reactive protein CRP
Procalcitonin PCT
Angiopoietin-like 3 Ang-like 3 Complement factor D Factor D
Interleukin 18 Binding Protein ILI8bp Combinations of biomarkers of the present invention includes any combination of 2, 3,4, 5, 6, 7, 8, 9, 10, 11, 12, 13, or all of the biomarkers listed in Table I
can be used. For instance, the number of possible combinations of a subset m of n proteins in Table 1 above is described in Feller, Intro to Probability Theory, Third Edition, volume 1, 1968, ed. J. Wiley, using the general formula:
T11!1(n)! (rn-n)!
In one embodiment of the invention, where n is 2 and in is 14, the number of combinations of protein markers selected from Table 1 is:
14! = 14x13x 12x 1 lx10x9x8x7x6x5x4x3x2x1 21(14-21! (2x1) (12x1 1 xl0x9x8x7x6x5x4x3x2x1) unique two-gene combinations. The measurement of the gene expression of each of these two-gene combinations, in an additive manner, can be used as described herein.
In another embodiment there are 14!/3M 4-3)! or 364 unique three-gene combinations and the measurement of each of these three-gene combinations, in an additive manner, can be used as described herein.
5.5 Biomarker Quantification Protein biomarkers to be quantified are often first isolated from a sample using techniques which are well known to those of skill in the art. Protein isolation methods can, for example, be such as those described in Harlow and Lane (Harlow, E. and Lane, D., Antibodies: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York (1988)).
Detection of quantity or level of the biomarkers in a sample can occur either directly in said sample, or upon further isolation or purification of extracted proteins using one or more techniques known in the art including density gradient centrifugation, ultra-centrifugation, concentration, dialysis, chromatography, precipitation, electrophoresis, flow preparation electrophoresis, selective banding and the like. Commercially available products for purification of proteins from samples, including blood, are also well known in the art including Qiagene's AllPrep DNA/RNA/Protein Mini Kit, and Molecular Research Centre's (MRCO) Tri-Reagent BD-RNA/DNA Protein Isolation Blood Derivative.

=
Protein biomarkers of a sample can also be differentiated upon purification or partial purification using such standard techniques such as a sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), potentially in combination with western blotting. Quantities of protein biomarkers can be determined using techniques known in the art. Useful ways to determine such levels include, but are not limited to, Western blots, protein microarrays, and Enzyme-Linked Immunosorbent Assays ("ELISA") and the like. A
number of different types of other useful assays that measure the presence of a protein biomarker are well known in the art. Immunoassays may be homogeneous, i.e.
performed in a single phase, or heterogeneous, where antigen or antibody is linked to an insoluble solid support upon which the assay is performed. Sandwich or competitive assays may be performed. The reaction steps may be performed simultaneously or sequentially.
Threshold assays may be performed, where a predetermined amount of analyte is removed from the sample using a capture reagent before the assay is performed, and only analyte levels of above the specified concentration are detected. Assay formats include, but are not limited to, for example, assays performed in test tubes, wells or on immunochromatographic test strips, as well as dipstick, lateral flow or migratory format immunoassays. Such examples are not intended to limit the potential means for determining the level of a protein biomarker in a sample.
Agents for detecting a a protein biomarker may utilize a complementary binding partner capable of binding to a protein of interest. A suitable complementary binding partner can include a nucleic acid aptamer, a peptide aptatner, a peptibody, a mimetic, a polyclonal antibody, a monoclonal antibody or any other protein or nucleic acid, or fragment thereof which is known to have specific interaction with the protein biomarker either in vivo or in vitro, or combinations thereof.
Complementary binding partners, including antibodies, can be conjugated to non-limiting materials such as magnetic compounds, paramagnetic compounds, other proteins such as avidin and/or biotin, nucleic acids, antibody fragments, or combinations thereof and/or can be disposed on an appropriate surfaces to allow detection including glass, polystyrene, polypropylene, polyethylene, dextran, nylon, amylases, natural and modified celluloses, polyacrylamides, gabbros, and magnetite NPV membrane, plastic, including a support intended to be used as a dipstick or a support useful for a microarray.
One or more complementary binding partners used for quantification of the protein biomarker can be operably linked (attached via either covalent or non-covalent methods) to a detectable label. Methods for linking said detectable label to a complementary binding partner is well known in the art (see, e.g., Wong, S. S., Chemistry of Protein Conjugation and Cross-Linking, CRC Press 1991; Burkhart etal., The Chemistry and Application of Amino Crosslinking Agents or Aminoplasts, John Wiley & Sons Inc., New York City, N.Y., 1999).
Useful labels can include, without limitation, fluorophores (e.g., fluorescein (FITC), phycoerythrin, rhodamine), chemical dyes, fluorescent dies or compounds that are radioactive, chemiluminescent, magnetic, paramagnetic, promagnetic, or enzymes that yield a product that may be colored, chemiluminescent, or magnetic. The signal is detectable by any suitable means, including spectroscopic, photochemical, biochemical, immunochemical, electrical, optical or chemical means. In certain cases, the signal is detectable by two or more means.
All protein biomarkers are easily purified from blood, and can be readily used to generate monoclonal and/or polyclonal antibodies using traditional techniques for antibody generation well known in the art. Monoclonal antibodies can be prepared, e.g., using hybridoma methods, such as those described by Kohler and Milstein, Nature, 256;495 (1975) or can be made by recombinant DNA methods (U.S. Pat. No. 4,816,567). See also Goding, Monoclonal Antibodies Principles and Practise, (New York: Academic Press, 1986), pp. 59-103. Kozbor, J. Immunol., 133:3001 (1984); Brodeur et al., Monoclonal Antibody Production Techniques and Applications (Marcel Dekker, Inc.: New York, 1987) pp. 51-63.
Monoclonal and/or polyclonal antibodies that have been used or are known to be available as potentially useful complementary binding partners for detecting the protein biornarkers are disclosed in Table 2 herein.
Table 2 fe.01.01!:Ninie-:': : .1:14t, =':;.,-'' -:--commercially ,--:
lip .:=,,t,', , . iin e.
,!..;,., ..:i; u.' .... ..... yttibol .,:: -A*iiiiable.'!.
4:: .= ; ...: ¨ ..,;:ii; . Antibosli!..,.
,..:.,..,2;:::!.:.-Refekenee-;.. . ..., - -,:i::;,;:i.,-7::!....,õ : ..::,2!7,i:::::
Tit, . = . :='-',:- .!'::;:!:6 Complement Abeam fragment C5a C5a ab I 1878 beam Angiopoietin-1 Ang-I eb8451 _ ________________ ___________________________ 1 _________ lAbcame Angiopoietin-2 Ang-2 lab8452 kDa interferon Abeam gamma-induced ab8098 protein IP-10 Soluble intercellular R&D Systems adhesion molecule- Mab720 sICAM-1 Vascular Abeam endothelial growth Ab46 i 54 factor A VEGF
Soluble vascular R&D Systems endothelial growth Mab321 factor receptor 1 sFlt-1 Chitinase-3-like Abeam protein 1 CI-113L1 Ab93034 Soluble triggering Abeam receptor expressed Ab93717 on myeloid cells-I sTREM-1 Abeam C-reactive protein CRP Ab76434 Abeam r rocalc iton in PCT Ab53897 R&D Systems = ngiopoietin-like 3 = ng-like 3 MAb38291 omplement factor R&D Systems Factor D Mab1824 Interleukin 18 beam inding Protein ILI 8bp JAb529l4 5.6 Use of Biomarkers and Biomarker Combinations 5 As taught herein, one or more biomarkers or biomarker combinations can be used to determine the likelihood of a test individual having, or not having a critical and/or life threatening response to illness. In one aspect, the test individual has been diagnosed or differentially diagnosed, prior to use of the biomarkers or biomarker combinations. In another aspect, the test individual has not been diagnosed or differentially diagnosed prior to 10 the use of the biomarkers or biotuarker combinations. In other aspects, the test individual has been diagnosed with one or more symptoms indicative of having an illness, but the source or cause of the illness remains unknown prior to the use of the biomarker or biomarker combinations.
In some embodiments, the biomarker and biomarker combinations determine that the test individual has an increased risk of having a critical and/or life threatening response.
In some embodiments, the biomarker and biomarker combinations determine that the test individual has a decreased risk of having a critical and/or life threatening response. In some embodiments, the biomarker and biomarker combinations determine that the test individual has at a significantly increased risk of having a critical and/or life threatening response. In some embodiments, the biomarker and biomarker combinations determine that the test individual has a significantly decreased risk of having a critical and/or life threatening response .The increased risk or decreased risk is in comparison to a control population. In some embodiments, the control population is a negative control population of individuals not having an increased risk of a critical and/or life threatening response to illness. In some embodiments, the control population is a positive control population of individuals having an increased risk of a critical and/or life threatening response to illness. In some embodiments, the control population is a population of individuals who have had the illness of the test individual and have not developed a critical or life threatening response. In some embodiments the control population is population of normal individuals. In some embodiments, the control population is an unbiased population of individuals with the same illness as the test individual. In some embodiments, the control population is a population of individuals who have had the illness and have developed a critical or life threatening response.
In order to determine the likelihood of an individual having a critical and/or life threatening response to an illness, the levels of one or more of the protein biomarkers of Table 1 in a sample are detecting and quantified and compared with the quantified control levels of said one or more protein biotnarkers in a control population.
For each individual protein biomarker, where the level of the protein biomarker in the test individual is significantly different (where by significantly different is meant a statistically significant difference) from the level of the protein biomarker in the control population, it aids in the determination that the test individual is likely to have a different response to a critical and/or life threatening response to illness than the control individual.
In some embodiments, the results from a single biomarker may be sufficient to determine that the test individual is at an increased or decreased risk of having a critical and/or life threatening response to illness. Whether a single biomarker is sufficient to determine that the test individual is at an increased or decreased risk of having a critical and/or life threatening response to illness will depend upon the desired sensitivity and/or specificity of the test results. In some embodiments, it will be sufficient that the sensitivity is greater than 51% and the specificity is greater than 51%. In other embodiments, the sensitivity of the test results must be greater than 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99% or must be 100%. In some embodiments the specificity of the test results must be greater than 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%
or must be 100%.
In seine embodiments, in order to achieve the desired sensitivity and/or specificity of the test results, two or more biomarkers, three or more biomarkers, four or more biomarker, five or more biomarkers, six or more biomarkers, seven or more biomarkers, eight or more biomarkers, nine or more biomarkers, ten or more biomarkers, 11 or more biomarkers, 12 or more biomarkers, 13 or more biomarkers, or all biomarkers must be used in combination.
In some embodiments, each of said two or more biomarkers, three or more biomarkers, four or more biomarker, five or more biomarkers, six or more biomarkers, seven or more biomarkers, eight or more biomarkers, nine or more biomarkers, ten or more biomarkers, II or more biomarkers, 12 or more biomarkers, 13 or more biomarkers, or all biomarkers are weighted equally to make a determination with respect to the status of a test individual.
In some embodiments. in order to achieve the desired sensitivity and/or specificity, each of said biomarkers in the combination may be weighted differently as determined by a classifier using at least two populations, wherein at least one population has been pre-determined to have a critical and/or life threatening response to an illness, and at least one population has been pre-determined to not have a critical and or life threatening response to an illness.
In some embodiments the classifier is built using logistic regression as the mathematical model. In other embodiments, the classification tree analysis is used.
5.7 Kits The present invention provides kits for measuring the levels of at least 1, at least 2, at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, at least

11, at least 12, at least 13 or any or all combinations of the protein biomarkers of the invention. Such kits comprise materials and reagents required for measuring the levels of such protein biomarkers. As such, the kits provide one or more complementary binding proteins to measure the level of said biomarkers of said combinations. In some embodiments the complementary binding proteins are monoclonal antibodies, and the kit includes antibodies which bind specifically to each of biomarkers to be measured. The kits may additional comprise one or more additional reagents employed in the various methods, such as (I) one or more labelled or non-labelled antibodies which can bind the complementary binding proteins in said kit (eg. Anti-mouse antibodies (1) labeling reagents ( (2) one or more buffer mediums, e.g., hybridization and washing buffers; (3) protein purification reagents; (4) signal generation and detection reagents, e.g., streptavidin-alkaline phosphatase conjugate, ehemifluorescent or ehemilumineseent substrate, and the like. In particular embodiments, the kits comprise prelabeled quality controlled protein for use as a control.
In some embodiments, an antibody based kit can comprise, for example: (1) at least one first antibody (which may or may not be attached to a support) which binds to a specific protein biomarker; (2) a second, different antibody which binds to either the protein biomarker, or the first antibody and is conjugated to a detectable label (e.g., a fluorescent label, radioactive isotope or enzyme). The antibody-based kits may also comprise beads for conducting an immunoprecipitation. Each component of the antibody-based kits is generally in its own suitable container. Thus, these kits generally comprise distinct containers suitable for each antibody. Further, the antibody-based kits may comprise instructions for performing the assay and methods for interpreting and analyzing the data resulting from the performance of the assay. In a specific embodiment, the kits contain instructions for determining the likelihood an individual is at an increased risk of a critical and/or life threatening response to illness.
5.8 Examples Example 1 Individual Biomarkers Predictive of Outcome in Pre-Diagnosed Malaria A retrospective case-control study was performed at Mulago Hospital in Kampala studying children with malaria as the illness. Children were enrolled between the ages of 6 months and 12 years old who presenting with clinical signs and symptoms of malaria wherein the diagnosis was confirmed by detecting the presence of?.
falciparzim infections by microscopic analysis were utilized. Children with co-morbidities such as sickle cell trait/disease, HIV co-infection or severe malnutrition were excluded.
Using plasma banked samples, various protein biomarkers were isolated and measured in the plasma from the approximately 100 Ugandan children where the diagnosis was confirmed as either cerebral malaria (CM) or severe malarial anemia (SMA) (both illness that can progress to life threatening disease). The levels of each of a selection of specific protein biomarkers was measured in the banked plasma samples and compared as between children who were known to have survived the malaria as compared with the levels of these protein biomarkers in children who died.
Plasma samples were isolated from whole blood after treatment with sodium citrate anticoagulant, and were stored at -20 C prior to testing. ELISAs were used to quantify the levels of various potential biomarkers including Ang-2, CRP, sTREM-1, 1P-10, sFlt-1, sICAM-1,and PCT, in said samples. ELISAs were performed in accordance with manufacturer's instructions with the following changes: assays were performed in a volume of 50 aL/well; plasma samples were incubated overnight at 4 C; and ELISAs were developed using Extravidin*- Alkaline Phosphatase (Sigma, 1:1000 dilution, 45 min incubation) followed by addition of p-Nitrophenyl phosphate substrate (Sigma) and optical density readings at 405 urn. Assays were developed with tetramethylbenzidine, stopped with H2SO4, and read at 450 nm. Samples with concentrations below the limit of detection were designated as twice the background level. Background signal was determined from blank wells included on each plate (assay buffer added instead of sample), and background optical density was subtracted from all samples and standards prior to analysis.
Samples with optical densities below the lowest detectable standard were assigned the value of that standard.
GraphPad Prism v4, SPSS v18, and MedCalc software were used for analysis.
For clinical and demographic variables, differences between groups were assessed using the Chi-square test (categorical variables) or the Kruskal-Wallis test with Dunn's multiple comparison post-hoc tests (continuous variables). The Mann-Whitney U test was used to compare biomarker levels between groups, and p values were corrected for multiple comparisons using Holm's correction.
Levels of protein biotnarkers were compared as between children who survived the malaria as compared with children who died from the malaria and are presented as dot plots with medians shown in Figure 1A. Figure I B demonstrates results on the same population for the biomarker sTREM-I, and the dotplot categorizes the individuals has having either survived or died. A Mann Whitney U test was performed for each comparison to determine the statistical significance of the difference as between the two populations of levels, and those biomarkers showing a statistically significant difference between the two populations is shown with a * (p<0.05) or ** (p<0.01) in Figure 1A, 1B. Within this small sample size, sTie-2 did not reach statistical significance. Nevertheless, given the dose interaction between sTie-2 (as the receptor to Ang-2), the fact that Ang-2 did show a statistically significant response, and given the differential trend seen for sTie-2 (despite not reaching statistical significance) we reasonably predict that this biomarker will demonstrate utility when tested with sample populations in greater numbers.
Receiver operating characteristic curves were generated using the non-parametric method of Delong et. al (DeLong ER, DeLong DM, Clarke-Pearson DL
(1988) Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 44:837-845). Data is shown for biomarkers sICAM-1, sFlt-1, Ang-2, PCT, IP-10, and sTREM-1 in Figure 2A. As would be understood the area under the ROC curve is indicative of the ability of each biomarker to differentiate between the likelihood of an individual dying and not dying. Shown in dashed reference lines is an ROC curve for a test which has no discriminatory ability. The area under the ROC curve is noted and its statistical significance as either * p<0.05 or ** p<0.01 shown.
In parenthesis is the 95% confidence intervals for the area under the curve. Figure 2B shows the ROC curve for parasitemia, which is currently relied upon to assess the individual's response to malaria.
Parasitemia predicts the quantitative content of parasites in the blood and is used as a measurement of parasite load in the organism and an indication of the degree of an active parasitic infection. As can be seen, each of the biomarkers noted is better at predicting death than the currently utilized index of parasitemia.
To evaluate the biomarkers further, the Youden index was used to obtain a cut-point for each biomarker, and clinical performance measures evaluated for these dichotomized biomarkers (Table 3). All parameters presented in Table 3 are presented with 95% confidence intervals shown in brackets. All cut points were determined using the Youden Index (J-max[sensitivy+specificity-1]). For each biomarker is shown the PLR, positive likelihood ratio, NLR the negative likelihood ratio, PPV, the positive predictive value and NPV, the negative predictive value. PPVs and NPVs were based on estimates that 5.7% of CM and SMA diagnosed patients at the Mulago hospital died of the malaria infection. sTREM-1 achieved the highest sensitivity (95.7%) but had low specificity (43.8%), while IP-10 predicted death with the highest overall accuracy (82.6%
sensitivity, 85% specificity).

Table 3 Clinical Performance of Biomarkers for Predicting Mortality Among Children with Severe Malaria Cut-point sensitivity Irecificity PI ,R NLR ippV(%) INPV
(%) (%) %) Ang-2 >5.6ng/m1 78.3 178.8 3.7 0.3 18.2 98,4 (56.3-92.5) 1(68.2-87.1) (2.9-4.7) (0.1-0.7) 5.8-38.7) (92.4-99.9) sICAM >645.3ng/m1 87.0 175.0 3.5 02 17.4 99.0 (66.4-97.2) $64.1-84.0) (2.8-4.3) (0.06-0.5) (5.9-35.9) (93.2-100) sF11-1 >1066.3pg/m1 82.6 157.5 1.9 0.3 10.5 98.2 (61.2-95.0) I 45.9-68.5) (1.5-2.5) (0.1-0.8) (3.4-23.1) (90.4-100) PCT >43.1ng/m1 56.5 182.5 3.2 0.5 16.3 96.9 (34.5-76.8) k 72.4-90.1) (2.2-4.7) (0.3-1.0) (3.8-39.5) (90.5-99.5) IP-10 >831.2pg/m1 82,6 185.0 5.5 0.2 25 98.8 (61.2-95.0) 05.3-92.0) (4.5-6.8) (0.07-0.6) (8.3-49.8) (93.4-100) sTREM-I1>289.9pg/m1 95.7 43.8 1.7 0.1 9.3 99.4 (78.1-99.9) 32.7-55.3) (1.3-2.2) (0.01-0.7) (3.3-19.6) (90.5-100) Example 2 Biomarker Combinations Predictive of Mortality in Pre-Diagnosed Malaria Data was obtained as described in Example 1. The use of biomarker combinations improved the ability to predict the likelihood of an individual's life threatening response in malaria. In this example, a modest number of deaths in the study precluded using multivariable logistic regression analysis to create classifiers with more than 2-3 independent variables (Harrell FE, Jr., Lee KL, Mark DB (1996) Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med 15:361-387). Therefore, as performed in other conditions, (Morrow DA, Braunwald E (2003) Future of biomarkers in acute coronary syndromes: moving toward a multimarker strategy. Circulation 108:250-252; Vinueza CA, Chauhan SP, Barker L, Hendrix NW, Scardo JA (2000) Predicting the success of a trial of labor with a simple scoring system.
J Reprod Med 45:332-336), six biomarkers were combined (Ang-2, sICAM-1, sFlt-1, PCT, IP-10 and TREM-I) into a single score. For each marker, one point was assigned if the measured value was greater than the corresponding cut-point, and zero points were assigned if it was lower. A cumulative "biomarker score" was calculated for each patient by summing the points for all six markers. No two dichotomized biomarkers were highly correlated (Spearman's rho <0.6; data not shown), suggesting that each biomarker would contribute unique information to the score since biomarkers which are not correlated indicate that the biomarkers each add new information as compared with single biomarkers alone.
Biomarker score was highly positively correlated with risk of death (data not shown; Spearman's rho=0.96, p=0.003). Scores were elevated among fatalities compared to survivors (median (interquartile range): 5 (4-6) and 1 (0-2.5), respectively, data not shown.
In a univariate logistic regression model, the biomarker score was a significant predictor of death with an odds ratio of 7.9 (95% Cl 4.6-54.4) (Table 4, Model I).
After adjustment to exclude parasitemia and age, which have been associated with malaria mortality as predictive factors, the score remained significant with an adjusted odds ratio of 7.8 (4.7-134) (Table 4, Model 2).
ROC curve analysis and cut-point determination were performed as above for various biomarker combinations to determine their utility in as predictive indicators of outcome of illness. Table 5 shows the data resulting from the biomarker combinations tested. All combinations demonstrated some utility as predictive indicators of outcome of illness. Additional combinations are shown in Table 6. All parameters in both Table 5 and Table 6 are presented with 95% Cls in parentheses.
Cut-points were determined using the Youden Index (J = max[sensitivity +
specificity ¨
I]). PLR indicates the positive likelihood ratio; NLR indicates the negative likelihood ratio; PPV is the positive predictive value; and NPV is the negative predictive value.
Using logistic regression on the six biomarker combination of Ang-2, sICAM- I, sFlt-1, PCT, 1P-10 and TREM-1, the AUC was 0.96 (0.90-0.99) (data not shown), and a score >4 was found to have a 95.7% sensitive and 88.8% specific for predicting death in the samples tested (Table 5, row I). For logistic regression, linearity of an independent variable with the log odds of the dependent was assessed by including a Box-Tidwell transformation into the model and ensuring that this term was not significant. Bootstrapping (1000 sample draws) was used to generate variance estimates for the cut point. Model goodness-of-fit was assessed by the Hosmer-Lemeshow test and calibration slope analysis (Steyerberg EW, Eijkemans MJ, Harrell FE, Jr., Habbema JD (2001) Prognostic modeling with logistic regression analysis: in search of a sensible strategy in small data sets. Med Decis Making 21:45-56.).
Positive and negative predictive values were calculated using the reported case fatality rate of 5.7% for microscopy-confirmed CM and SMA cases. (Hosmer DW, Leirieshow S. Applied Logistic Regression. 2nd ed. New York: John Wiley & Sons, Inc, 2000).PPVs and NPVs were based on estimates that 5.7% of CM and SMA patients at the Mulago hospital where samples were obtained die of the malaria infection. While the positive predictive value for the six biomarker combination was low (33.9%) given a fatality rate of 5.7%, the negative predictive value (NPV) was 99.7%, indicating that a child with a score <3 will likely respond well to standard treatment protocols.

Table 4. Association of biomarker score with outcome among children with severe malaria: logistic regression.' Hosmer-Lemeshow test Variable b (95% Cl) SE Wald df p value OR (95% Cl) Chi square df p value Model 11) Biomarker score 2.1 (L5-4.0) 2.3 18.6 1 0.00! 7.9(4.6-54.4) 3.3 5 0.66 Model 2c Biornarker scored 2.1 (1.6-4.9) 21.5 18.2 I 0.001 7.8 (4.7-134) 11 8 1.0 Log parasitemiac 0.050 ((-1.1)-1.3) 2.8 0.010 1 0.91 1.1 (0.35-3.6) Age 0.053 ((-0.61)-1.2) 8.5 0.052 I 0.89 1.1 (0.55-3.3) 'The reference category was "survival." b1'seudo-R2 (Cox & Snell) 0.473 and calibration slope 0.98. 93seudo-R2 (Cox & Snell) 0.474 and calibration slope 1Ø dBiomarker score and log parasitemia had a significant but low correlation (Spearman's rho 0.292, p<0.01).
Varasitemia was log-transformed in order to achieve linearity with the log-odds of the dependent variable. SE, standard error; OR, odds ratio.

=
Table 5. Clinical performance of biornarker combinations tbr predicting mortality among children with severe malaria.' Biomarker combination Numberdd of hreshold Sensitivity Specificity - rP1)' V -FNPV
i ---n ivi uals utilized positives based on (%) (%) in generating the 1ROC curves) data (n).
' _______________________________________________________________________________ ________ [P-10, sICAM1 104 6/2 77.3 96.6 85 94.4 0 IP-10, sICAM1 98 (exclude non- /2 IV
--.1 CIWS MA fatal) 93.8 96.3 83.3 , 8.8 0, ko 0.
w ANG-2, IPI 0, sICAM1 104 '/3 86.4 87.5 63.3 96.3 w N., ANG-2, IP10, sICAM I 98 (exclude non- r/3 N., CM/SMA fatal) 93.8 86.6 57.7 98.6 0 N., ANG-2, IPI 0, CHI3L1 77 /3 93.8 82.0 57.7 98.0 IV
--.1 ANG-2, IP10, sTREM1 77 /3 93.8 85.2 62.5 98.1 ANG-2, sICAMI, CH131,1 77 /3 93.8 93.4 78.9 98.3 ANG-2, sICAM1, sTREM1 77 /3 93.8 88.5 68.2 98.2 ANG-2, CI-113L1, sTREM I 77 /3 81.3 85.2 59.1 94.5 .
[PIO, sICAM1, CIII3L1 77 t3 93.8 88.5 68.2 98.2 IP10, sICAM I, sTREM I 77 /3 93.8 86.9 65.2 98.1 , , ) sICAM I , CHI31, I , sTREM I 77 2/3 93.8 86.9 65.2 98.1 ANG-2, IPIO, sICAM I, 0113L1 77 2/4 100.0 80.3 57.1 100.0 ' _______________________________________________________________________________ _________ _I
ANG-2, 1P10, sICAM I, sTREM I 77 2/4 100.0 78.7 55.2 100.0 ANG-2, sICAM1, CHI3L I , sTREM I 77 2/4 100.0 .82.0 59.3 100.0 I
IPIO, sICAM I, CH131,1, sTREM I 77 2/4 100.0 77.0 53.3 '100.0 ANG-2, IPIO, CHI31,1, sTREM1 77 2/4 100.0 73.8 50.0 100.0 0 ANG-2, IP I 0, sICAM I, CI-113L1 77 3/4 87.5 93.4 77,8 96.6 0 "
--.1 1.0 ANG-2, IPIO, sICAM1, sTREM1 77 3/4 87.5 95.1 82.4 196.7 0.
w w ANG-2, slCAM I, CI-13L1, sTREM1 77 3/4 81.3 93.4 76.5 i 5.0 "

1-.
N., IP I 0, sICAM1, CH131,1, sTREM I 77 3/4 87.5 95.1 82.4 961 0 N., I
ANG-2, IP I 0, CHBL I , sTREM1 '77 3/4 81.3 93.4 76.5 95.0 IV
--.1 ANG-2, [P10. sICAM I, C1-113L 1 , 77 3/5 sTREM1 ______________________________________________________________________ _____100 91.8 76.2 1100 _______________________________________________________________________________ ________ Table 6. Clinical performance of selected biomarker combinations for predicting mortality among children with severe rnalaria.a Combination Cut-pointb Sensitivity (%) Specificity (%) PLRc NLR PPV (%)d NPV (%) (Ang-2, 0.05 33.9 sICAM-1, sFlt- 95.7(78.1-99.9) 88.8 (79.7-94.7) 8.5 (7.6-9.6) 99.7 (95.2-100) (0.007-0A) (12.8-61.3) I. PCT, IP- I 0) Ang-2. PCT, >1 0.1 32.9 91.3 (72.0-98.9) 88.8 (79.7-94.7) 8.1 (7.0-9.4) 99.4 (94.7-100) sICA M-1 ¨ (0.02-0.4) (12.1-60.3) Ang-2, 1P-10, _ 0.1 28.6 0 PCT 91.3 (72.0-98.9) 86.3 (76.7-92.9) 6.6 (5.7-7.7) (0.02-0.4) (10.2-54.4) 99.4 (94.6-1 00) PCT, IP-10, 0.1 22.7 91.3 (72.0-98.9) 81.3 (71.0-89.1) 4.9 (4.1-5.7) 99.4 (94.2-100) sTRE M-1 (0.03-0.4) (8.1-44.8) Example 3¨Use of Classification Tree Analysis as an Alternative Classifier Predictive of Mortality in Pre-Diagnosed Malaria To explore other synergistic combinatorial strategies, wherein weighting of each biomarker may vary, classification tree analysis was used, which selects and organizes independent variables into a decision tree that optimally predicts the dependent measure.
Initially, a model based on IP-10 and sTREM-1 was generated with 43.5%
sensitivity and 100% specificity for predicting mortality (Figure 3). Since in some instances high sensitivity would be of particular importance, the analysis assigning the cost of misclassifying a death as a survivor was weighted as being 10 times greater than the cost of misclassifS,ing a survivor as a death. A model based on IP-10, Ang-2, and sICAM-1 was generated with 100%

sensitivity and 92.5% specificity for predicting outcome (cross-validated misclassification rate 15.4%, standard error 4.9%). In summary, combining dichotomized biomarkers using a scoring system or a classification tree predicted severe malaria mortality in our patient population with high accuracy.
Example 4 Individual Biomarkers and Biomarker Combinations Predictive of Patients Developing Toxic Shock Syndrome in Patients with Invasive S. pyogenes Disease A prospective, population-based surveillance for invasive group A
streptococcal disease was undertaken in Ontario, Canada via mandatory laboratory reporting of S_ pyogenes isolates from normally sterile sites and thirty-seven patients, enrolled between 1999 and 2009, were included in the study. Informed consent was obtained to collect bacterial isolates and plasma samples, as well as detailed clinical data from interviews with the attending physicians and patient chart review. Patients were considered to have S.
pyogenes infections which resulted in streptococcal toxic shock syndrome (STSS) (a critical and/or life threatening form of an S. pyogenes infection) if they met the current consensus of indicator symptoms including: hypotension in combination with at least two of coagulopathy, acute renal failure, elevated serum aminotransferases, acute respiratory distress syndrome (ARDS), rash, or necrotizing fasciitis. Of the 37 patients, 16 were considered to have invasive streptococcal infection and toxic shock (STSS), while 21 were determined to have invasive streptococcal infection alone (no STSS). The underlying source of the infection was similar between the two groups, with the majority of patients in both groups having skin and soft tissue infections (7 patients (44%) with STSS and 12 patients (57%) with invasive streptococcal infection alone). Presenting group A streptococcal infections in the remaining patients included respiratory tract infections, bacteremia without an identified source, post-partum infection, and peritonitis, and did not differ significantly between the groups. The two groups were significantly different only in the symptomatic diagnostic criteria for SISS;
hypotension was present in 100% of patients with STSS and 33% of patients without (P <
0.0001). Five patients with invasive infection and STSS died as compared to one patient with invasive infection alone (31% versus 5%, P = 0.06).
Acute phase plasma samples were collected upon study enrollment and stored at minus 70 C until use. Plasma concentrations of angiopoietins-1 and -2 were measured by ELISA (R&D Systems, Minneapolis MN) according to the manufacturer's instructions. The upper and lower limits of detection for the assays were 10,000 pg/m1... and 9.77 pern1_, for Ang-1 and 2520 pg/mL and 2A6 pg/ird.., for Ang-2, respectively. Samples were diluted in assay diluent (1:20 for Ang-1 and 1:4 for Ang-2) to fall within the range of the standard curves.
Angiopoietin dysregulation (a correlated decrease in Ang-1 levels and an increase in Ang-2 levels) was associated with an increased likelihood of the individual having the invasive group A streptococcal disease with STSS as compared with individuals having invasive group A streptococcal disease without STSS (Figure 4A and Figure 4B).
The median plasma concentration of Ang-1 was lower during the acute phase of illness in patients pre-diagnosed with invasive infection and STSS than in those pre-diagnosed with invasive streptococcal infection alone (13,915 pg/mL vs. 29,084 pg/mL), while the median plasma concentration of Aug-2 was higher (5752 pg/mL vs. 1337 pg/mL). As a result, the normally low Ang-2:Ang-1 ratio was significantly higher amongst patients with invasive infection and STSS as compared to those with invasive streptococcal infection alone (0.437 versus 0.048, P
<0.05).
Receiver operating characteristic (ROC) curves were generated for Ang-1, Ang-2, and the Ang-2:Ang-1 ratio, and the area under the ROC curves indicated that the degree of magnitude of Ang-1/2 dysregulation accurately differentiated those individuals with STSS from those without STSS (Figure 4B). Although the ROC curve for plasma Ang-1 concentration did not differ significantly from chance (AUC: 0.683, P =
0.07), the ROC
curves for plasma Ang-2 (AUC: 0.759, P -= 0.009) and for the Ang-2:Ang-1 ratio (AUC:
0.791, P = 0.003) revealed that both discriminated between patients with STSS
and those with invasive streptococcal infection alone (no STSS) and it is anticipated that the ROC curve LRR Patent Law Office -228 Degrassi Street , -,t,t A t µ,0 I Al& /
t = ' =rfr"14 6, Toronto, Ontario LRR Patent Law Office Industry Industrie NM/J ....
Y/M/D
Canada Canada Phone: 416-465-6565 Fax: 416-465 6767 CIPO OPIC
cool 824452 E-mail. kapaport@Irrpatentlaw.corn 228 Degrassi Street Toronto, Ontario Fax Transmittal Form Attention: Formalities Phone:
Fax: 819-953-2476 To: Commissioner of Patents Your Name: Lesley Rapaport Phone: 416-465-6565 Fax: 416-465-6767 E-mail: Irapaport@Irrpatentlaw.com From: LRR Patent Law Office Date Sent:
fbwizid 219 Number of pages inclueing cover page:
(A(/-0Y-2- ) Ca if,/
Urgent For Review Please ComMent Please Reply Confidential Message:
041 (0/ii 46,0 for plasma Ang-1 would also be discriminatory upon an increased sample size since the the ROC curve for plasma Aug-1 concentration trended despite not reaching statistical difference (AUC: 0.683,P = 0.07).
Example 5 Individual Biomarkers and Biomarker Combinations Predictive of Response in Patients Having Group A Streptococcal Disease Using the samples and methods as outlined in Example 4, we further measured the biomarkers Aug-I, Ang-2 and the ratio of Ang-I/Ang-2 as the patients convalesced to demonstrate the potential for the biomarkers to function as indicators of response to treatment. Ang-1/2 dysregulation was seen to resolve consistent with convalescence in both groups of patients (Figure 4A). In the cohort of patients with STSS, the median plasma concentration of Ang-1 rose from 13,519 pg/mL to 21,115 pg/mL, the median plasma concentration orAng-2 decreased fell from 5752 pg/mL to 378 pg/mL (P <0.01), and the median Ang-2:Ang-1 ratio fell from 0.437 to 0.019 (P <0.05).
Furthermore, in individual patients with STSS, the matched acute and convalescent plasma Ang-2 concentrations and the Ang-2:Ang-1 ratios also differed significantly (Figure 5) The same pattern was observed in the cohort of patients with invasive streptococcal disease without STSS, the changes in Ang-1/2 concentrations although the changes were more modest. The median plasma concentration of Aug-1 in this group increased from 29,084 pg/mL to 31,743 pg/mL, while the Ang-2 concentration declined from 1337 pg/mL to 535 pg/mL, and the Ang-2:Ang-1 ratio decreased from 0.048 to 0.027 Example 6 Individual Biomarkers and Biomarker Combinations Predictive of Outcome in Pre-Diagnosed Sepsis A multicenter retrospective analysis was performed on prospectively collected biological and clinical data so as to identify molecular markers demonstrating an increased likelihood of patients dying from severe sepsis. Samples were collected from three tertiary hospital intensive care units (ICU) associated with Hamilton General Hospital in Hamilton, Canada.
Seventy patients with severe sepsis enrolled within 24-hours of admission to the ICU and were followed until day 28, discharge or death. Clinical data and plasma samples were available on admission for all patients and daily for 1 week, then weekly thereafter for 43 of the 70 patients.

Patients were diagnosed as having severe sepsis if they met the modified American College of Chest Physicians /Society of Critical Care Medicine criteria for sepsis known in the art (Bernard GR, Vincent J-L, Laterre P-F, et al. Efficacy and safety of recombinant human activated protein C for severe sepsis. N Engl J Med 2001;344(10):699-709; Bone RC, Sibbald WJ, Sprung CL. The ACCP-SCCM consensus conference on sepsis and organ failure. Chest I 992;101(6):1481-1483.) Patients were included if they had known or suspected infection as well as at least three of four modified SIRS criteria and at least one of five criteria for organ dysfunction.
Venous blood (4.5 ml) collected from indwelling catheters was transferred into 15 polypropylene tubes containing 0.5 ml of 0.105 M buffered trisodium citrate (pH
5.4) and 100 pi of 1 M benzamidine HO and centrifuged at 1,500 g for 10 min (20 C).
Plasma for analysis was stored in aliquots at -80oC. Commercial enzyme-linked immunoassays (ELISAs) were used to measure levels of biomarkers. Ang-1 and Ang-(R&D Systems, Minneapolis, MN, USA) were measured on available samples from days I to 7, 14, and 28. ESEL (R&D Systems, Minneapolis, MN, USA), sICAM-1 (R&D Systems, Minneapolis, MN, USA) and vVvT (antibody: Dako, Carpinteria, CA, USA;
standard:
American Diagnostica, Stamford, CT, USA), levels were measured on days 1 and 3. All standards, controls and test samples were assayed in duplicate and averaged prior to interpretation. Concentrations were interpolated from four parameter logistic fit curves generated using a standard curve of recombinant human proteins.
It was determined that patients with low Ang-1 plasma levels (<5.5 ng/mL) at admission were less likely to survive than those with high Ang-1 levels (> 5.6 ng/m1; relative risk 0.49 [95% Cl: 0.25 ¨ 0.98j, p=0.046 (Figure 6A).
Ang-1 levels <5.5 ng/mL also remained a significant predictor of mortality at 28 days in a multivariate logistic regression model (adjusted odds ratio 0.282 [95%
confidence interval (CI): 0.086-0.931, p=0.037) using known clinical indicators of increased risk of mortality. Age is a known risk factor leading to increased likelihood of death from sepsis. Similarly Multiorgan Dysfunction (MOD) score exists as the current method of measuring and quantifying organ disfunction, either as a risk factor for death, a measure of severity of illness, or a measure of increased risk for morbidity over time.
The multivariate logistic regression model used age (p=0.008) and MOD score (p-0.014) as additional clinical biomarkers, suggesting that Ang-1 provides independent prognostic information above and beyond age and MOD scores alone.

This finding is supported by receiver operating characteristic (ROC) curve analysis ( Figure 6B) illustrating the apparent added sensitivity and specificity in predicting 28-day mortality when comparing plasma Ang-1 levels (area under the ROC curve (AUROC): 0.62 [95% CI: 0.50-0.76D, MOD score (AUROC: 0.64 [95% CI: 0.51-0.77]) or age (AUROC:
0.68 [95% Cl: 0.55-0.80]) with the combination of the three variables (AUROC:
0.79 [95%
Cl: 0.67-0.901).
Example 7 Individual Biomarkers and Biomarker Combinations As Early Predictors of Risk of Mortality in Patients with Sepsis As noted, the current standard for determining an individuals increased likelihood of death from sepsis is the Multiorgan Dysfunction (MOD) score.
Using samples and methods as described in Example 6, the level of Ang-2 was measured and correlated with the MOD score across the population of individuals tested. As noted in Figure 7A, the level of Ang-2 correlated (as noted on the y axes in ng/m1) when compared with the MOD score (as noted on the x axis) as a predictor of mortality, with a statistical significance of p<0.000I
as tested using as a single biomarker was demonstrated. The ability of the Ang-2 levels to act as an earlier predictor of mortality was analyzed by similarly comparing the level of Ang-2 (ng/m1) taken from patients one day prior to the evaluation of the patient as determined by MOD score. As can be seen in Figure 78, Ang-2 levels measured on day x predicted the clinical condition on the next hospital day (i.e. day x+1). There was a strong statistical correlation (P<0.0001) between the Ang-2 levels performed on day x compared to the MOD
score on the next hospital day (day x+1), indicating Ang-2 is an earlier indicator of disease progression and risk of mortality than the current standard of the MOD score.
Example 8 Individual Biomarkers and Biomarker Combinations Predictive of Patients of Having Hemolytic Uremic Syndrome as a Result of an E. Cob Infection A population-based surveillance study for E. coli 0157:H7 infection in children less than 10 years of age was undertaken in Washington, Oregon, Idaho, and Wyoming through mandatory laboratory reporting of positive stool cultures.
Seventy-eight children, enrolled between 1998 and 2005, from whom a positive stool culture was obtained within the first 7 days of illness were included for this analysis. Phlebotomy was conducted at enrollment and as clinically indicated thereafter. HUS was diagnosed as hemolytic anemia (a hematocrit <30% with evidence of schistocytes on peripheral blood film), thrombocytopenia (platelet count < 150 000/mm3), and renal insufficiency (serum creatinine above the age-adjusted upper limit of normal); participants who had not met these criteria by day 14 of illness were considered to have had uncomplicated infection..
84 serum samples were tested: 26 from patients on the day of diagnosis of HUS, 8 from patients who would subsequently be diagnosed with HUS but had not yet met diagnostic criteria (pre-HUS), and 50 from patients with uncomplicated infection. Six patients had samples taken both prior to (pre-HUS) and on the day of HUS
diagnosis.
Serum samples were stored in aliquots at -80 C until use. To measure angiopoietin levels in cell culture supernatant, HMVEC were grown to confluence in complete medium in 6-well plates. Complete medium was replaced with basal medium lacking serum and growth factors on the day of toxin treatment. Shiga toxin or vehicle was added 4 hours later, and aliquots of were taken at 24 hours following toxin addition, centrifuged to remove dead cells, and likewise stored at -80 C until use.
Serum and supernatant concentrations of Ang-1 and Ang-2 were measured by ELISA (R8c1:;01) Systems, Minneapolis MN) as per the manufacturer's instructions. The technical upper limits of detection were 10,000 pg/mL for Ang-1 and 2520 pg/inL for Ang-2, yielding effective upper limits of detection of 200,000 pg/ml, and 10,080 pg/mL, respectively, for the dilutions employed in the assay. Lower limits of detection for the assay were 9.77 pg/mL for Ang-1 and 2.46 pg/mL for Ang-2.
Angiopoietin dysregulation (decreased Aug-1 and increased Ang-2) was found to be associated with illness severity. The median serum Ang-1 concentration in patients with uncomplicated infection was significantly higher than in those patients with HUS (77, 357 pg/mL [interquartile range (IQR): 53, 437 - 114, 889 pg/mL] versus 10, 622 pg/mL [IQR:
3464 - 43, 523 pg/mL]), P <0.001 (Figure 8A). Conversely, the median serum Ang-concentration was significantly lower in those with uncomplicated infection than in those with HUS (1140 [IQR: 845 - 1492 pg/mL} versus 1959 pg/mL [IQR: 1057 -pg/mL]), P <0.05. Finally, the Ang-2:Ang-1 ratio was 0.014 (IQR: 0.011 -0.023) in patients with uncomplicated infection, and more than 10-fold higher, at 0.18 (IQR).
In addition, the serum Ang-I concentration at the time of presentation to hospital effectively discriminated between two populations of clinically indistinguishable children: 1) those with uncomplicated hemorrhagic colitis and 2) those with hemorrhagic colitis who would eventually develop HUS (Area under the Receiver operating characteristic (ROC) curve [AUC): 0.785, 95% confidence interval (Cl): 0.641 - 0.923; P =
0.01) (Figure 8B).
The serum Ang- I and Ang-2 concentrations reported here for children with uncomplicated infection are comparable to those found in the serum of healthy children and adults, and are in keeping with the clinical observation that there is little if any endothelial activation present in these patients. In contrast, the relative deficit of Ang-1 and excess of Ang-2 found in children with HUS is in keeping with what is anticipated to be significant endothelial cell activation in these patients.
Example 9 Individual Diomarker of Outcome in Pre-Diagnosed Malaria and Use in Conjunction with Other Clinical Indicators of Outcome A retrospective case-control study was performed for children presenting with fever to the Queen Elizabeth Centre Hospital in Blantyre, Malawi. Children were between 6 months and 14 years of age and recruited between the years 1997 and 2009. EDTA
Plasma samples were obtained subsequent to obtaining informed consent. Children were characterized based on their status with respect to Cerebral malaria (CM) and also based on retinal indicators such as hemorrhages, retinal whitening, or vessel abnormalities. EDTA
Proteins isolated from Plasma samples were subject to ELISAs to quantify the levels of various potential biomarkers including Ang-2, Ang-1, and sTie-2.
Comparisons of continuous variables were performed using the Mann-Whitney U test and Spearman rank correlation coefficient. Comparisons of proportions were performed using the Person chi-square test, linear by linear association, or Fisher's exact test.
Odds rations (ORs_ were calculated using Pearson chi-square or logistic regression models to adjust for covariates. Bonferroni adjustmens were used to account for multiple comparisons.
Logistic regression and CRT analysis was used to generate prognostic models using routine clinical parameters in combination with the protein biomarkers. A
clinically predictive model of mortality was generated using solely the clinical parameters readily available (Age, BCS, respiratory distress, severe anemia), and probabilities from this clinical model were used to generate a c-index (equivalent to the area under the receiver operating characteristic curves) of 0.73 (95% confidence interval I_Cfj, 0.65-0.79) (data not shown).

Using these clinical model as a foundation, biomarker tests, either individually or in combination, were added to determine whether the biomarkers would significantly improve the predictive accuracy of the clinical parameters model alone. When the clinical model was combined with all three biomarkers Ang- I, Ang-2 and sTie-2, the resulting model had a c-index of 0.79 (95% confidence interval [C1j, 0.72-0.84) which was significantly better than the clinical model alone (p=0.03) (data not shown).
Example 10 Diagnosis of a Test Individual using Fliomarker Combination Predictive of a Critical and/or a Life Threatening Response.
Classifiers of the invention are generated using the detected levels of protein biomarkers Ang-1, Ang-2, !PIO and CHI3L1 in a population of individuals who demonstrate a critical and/or life threatening response to illness as compared with the detected levels of protein biomarkers Ang-2, IPI 0 and CHI3LI in a control population of individuals who are normal. Logistic regression is applied to differentiate the two populations and generates an equation which has a sensitivity of 90% and a specificity of 95%.
Levels of protein biomarkers Ang-2. IP 10 and C1-113L1 are determined using a standard EL1SA test on a serum sample from a test individual who may potentially have been exposed to an &coif infection, but has not yet been diagnosed with an E. con infection. In accordance with the logistic regression equation generated from the classifier as described, the test individual is classified as either having or not having a critical and or life threatening response to illness.

Claims

What is claimed is:

1. A composition comprising a collection of two or more antibodies and a suitable buffer, said composition capable of selectively binding to at least two protein biomarkers from a sample isolated from a test individual , wherein the protein biomarkers are: complement fragment C5a (C5a), angiopoietin-1 (Ang-1), angiopoietin-2 (Ang-2), 10 kDa interferon gamma-induced protein (IP-10), soluble tyrosine kinase with immunoglobulin-like loop and epidermal growth factor domain-2 (sTie-2), soluble intercellular adhesion molecule-1 (sICAM-1), vascular endothelial growth factor A (VEGF), soluble vascular endothelial growth factor receptor 1 (sFlt-1), Chitinase-3-like protein I
(CHI3L1), soluble triggering receptor expressed on myeloid cells-1 (sTREM-1), C-reactive protein (CRP), procalcitonin (PCT), angiopoietin-like 3 (Ang-like 3), complement factor D
(Factor D), or interleukin 18 binding protein (IL18bp) as set out in Table 1, and wherein the composition is used to quantify the level of said protein biomarkers in said sample.

2. The composition of claim 1, wherein the composition comprises a collection of at least three or more antibodies and is capable of selectively hybridizing to at least three protein biomarkers from the protein biomarkers: complement fragment C5a (C5a), angiopoietin-1 (Ang-1), angiopoietin-2 (Ang-2), 10 kDa interferon gamma-induced protein (1P-10), soluble tyrosine kinase with immunoglobulin-like loop and epidermal growth factor domain-2 (sTie-2), soluble intercellular adhesion molecule-1 (sICAM-1), vascular endothelial growth factor A (VEGF), soluble vascular endothelial growth factor receptor I
(sFlt-1), Chitinase-3-like protein 1 (CHI3L1), soluble triggering receptor expressed on myeloid cells-1 (sTREM-1), C-reactive protein (CRP), procalcitonin (PCT), angiopoietin-like 3 (Ang-like 3), complement factor D (Factor D), or interleukin 18 binding protein (IL18bp)as set out in Table 1.

3. The composition of claim 1 or 2, wherein said sample is a whole blood sample.

4. The composition of claim 1 or 2, wherein said sample is a serum sample.

5. The composition of claim 1 or 2, wherein said sample is a plasma sample.

6. The composition of claim 1, wherein the composition is capable of selectively binding to at least one protein biomarker which is: complement fragment C5a (C5a), vascular endothelial growth factor A (VEGF), soluble vascular endothelial growth factor receptor 1 (sFlt-1), Chitinase-3-like protein 1 (CHI3L1), C-reactive protein (CRP), angiopoietin-like 3 (Ang-like 3), complement factor D (Factor D), or interleukin 18 binding protein (IL18bp).

7. The composition of claim 6 wherein said sample is a whole blood sample.

8. The composition of claim 6, wherein said sample is a serum sample.

9. The composition of claim 6, wherein said sample is a plasma sample.

10. A composition comprising a collection of two or more antibodies and a suitable buffer, said composition capable of selectively binding to two or more proteins biomarkers from a sample isolated from a test individual , wherein the protein biomarkers are: complement fragment C5a (C5a), angiopoietin-1 (Ang-1), angiopoietin-2 (Ang-2), 10 kDa interferon gamma-induced protein (IP-10), soluble tyrosine kinase with immunoglobulin-like loop and epidermal growth factor domain-2 (sTie-2), soluble intercellular adhesion molecule-1 (s1CAM-1), vascular endothelial growth factor A (VEGF), soluble vascular endothelial growth factor receptor 1 (sFlt-1), Chitinase-3-like protein 1 (CHI3L1), soluble triggering receptor expressed on myeloid cells-1 (sTREM-1), C-reactive protein (CRP), procalcitonin (PCT), angiopoietin-like 3 (Ang-like 3), complement factor D
(Factor D), or interleukin 18 binding protein (IL18bp) as set out in Table 1, and wherein the composition is used to (i)detect and quantify a level of each of said two or more protein biomarker in said sample, (ii) compare said quantified levels of said protein biomarkers in said test individual to control levels of said protein biomarkers from a control population (iii) determining the presence of differential levels for each biomarker in the comparison of step (ii) so as to make a determination that said individual is at a significantly increased risk of a critical and/or life threatening response as compared with the control population.

11. The composition of claim 10, wherein said detecting and quantifying of step (i) utilizes one or more devices to transform the sample into data indicative of the levels of each of said two or more protein biomarkers which can be used to compare to the control population.

12. The composition of claim 11, wherein said one or more devices is an an enzyme linked immunoassay which is utilized so as to transform the sample into data.

13. The composition of claim 10, wherein said individual is subjected to a treatment protocol on the basis of the determination made in step (iii).

14. The composition of claim 10, wherein said control population is an unbiased population of individuals having the illness.

15. The composition of claim 10, wherein said control population is a population of individuals having the illness and not developing a critical and/or life threatening response to the illness.

16. The composition of claim 10, wherein said control population is a population of individuals who are normal.

17. The composition of claim 10, wherein said control population is a population of individuals wherein the majority of members of the control population do not have the illness.

18. The composition of claim 13, wherein said control population is an unbiased population of individuals having the illness.

19. The composition of claim 13, wherein said control population is a population of individuals having the illness and not developing a critical and/or life threatening response to the illness.

20. The composition of claim 13, wherein said control population is a population of individuals who are normal.

21. The composition of claim 13, wherein said control population is a population of individuals wherein the majority of members of the control population do not have the illness.

22. A method of determining the likelihood of an individual having a critical and/or life threatening response to an illness, said method comprising (i)detecting and quantifying a level of each of two or more protein biomarkers in a sample, wherein said protein biomarkers are: complement fragment C5a (C5a), angiopoietin-1 (Ang-1), angiopoietin-2 (Ang-2), 10 kDa interferon gamma-induced protein (IP-10), soluble tyrosine kinase with immunoglobulin-like loop and epidermal growth factor domain-2 (sTie-2), soluble intercellular adhesion molecule-I (s1CAM-1), vascular endothelial growth factor A
(VEGF), soluble vascular endothelial growth factor receptor I (sFlt-1), Chitinase-3-like protein 1 (CHI3L1), soluble triggering receptor expressed on myeloid cells-1 (sTREM-1), C-reactive protein (CRP), procalcitonin (PCT), angiopoietin-like 3 (Ang-like 3), complement factor D (Factor D), or interleukin 18 binding protein (IL18bp) as set out in Table 1 (ii) comparing said quantified levels of said protein biomarkers to control levels of said protein biomarkers from a control population (iii) determining the presence of differential levels for each biomarker in the comparison of step (ii) so as to make a determination that said individual is at an increased risk of having a critical and/or life threatening response to said illness as compared with the control population.

23. The method of claim 22, wherein the determination is made that the individual is at a significant increased risk of having a critical and/or life threatening response to said illness as compared with the control population.

24. The method of claim 22, wherein said detecting and quantifying of step (i) utilizes one or more devices to transform the sample into data indicative of the levels of each of said two or more protein biomarkers which can be used to compare to the control population.

25. The method of claim 24, wherein said one or more devices is an enzyme linked immunoassay which is utilized so as to transform the sample into data.

26. The method of claim 22, wherein said individual is subjected to a treatment protocol on the basis of the determination made in step (iii).

27. The method of claim 22, wherein said control population is an unbiased population of individuals having the illness.

28. The method of claim 22, wherein said control population is a population of individuals having the illness and not developing a critical and/or life threatening response to the illness.

29. The method of claim 22, wherein said control population is a population of individuals who are normal.

30. The method of claim 22, wherein said control population is a population of individuals wherein the majority of members of the control population do not have the illness.

31. The method of claim 26, wherein said control population is an unbiased population of individuals having the illness.

32. The method of claim 26, wherein said control population is a population of individuals having the illness and not developing a critical and/or life threatening response to the illness.

33. The method of claim 26, wherein said control population is a population of individuals who are normal.

34. The method of claim 26, wherein said control population is a population of individuals wherein the majority of members of the control population do not have the illness.

35. A method of determining the likelihood of an individual having a critical and/or life threatening response to an illness, said method comprising (i)detecting and quantifying a level of each of two or more protein biomarkers in a sample, wherein said protein biomarkers are: complement fragment C5a (C5a), angiopoietin-1 (Ang-1), angiopoietin-2 (Ang-2), 10 kDa interferon gamma-induced protein (IP-10), soluble tyrosine kinase with immunoglobulin-like loop and epidermal growth factor domain-2 (sTie-2), soluble intercellular adhesion molecule-1 (sICAM-1), vascular endothelial growth factor A
(VEGF), soluble vascular endothelial growth factor receptor 1 (sFlt-1), Chitinase-3-like protein 1 (CHI3L1), soluble triggering receptor expressed on myeloid cells-1 (sTREM-1), C-reactive protein (CRP), procalcitonin (PCT), angiopoietin-like 3 (Ang-like 3), complement factor D (Factor D), or interleukin 18 binding protein (IL18bp) as set out in Table 1, (ii) utilizing the quantified levels of each said protein biomarker from said sample in a classifier (iii) making a determination as to whether said individual is at a significantly increased risk of a life threatening response as compared with the control population.

36. The method of claim 34, wherein said detecting and quantifying the level of each protein biomarker in the sample utilizes one or more devices to tranform the sample into data which can be used in said classifier.

37 The method of claim 36, wherein said one or more devices is an an enzyme linked immunoassay which is utilized so as to transform the sample into data.

38 The method of claim 35, wherein said individual is subjected to a treatment protocol on the basis of the determination made in step (iii).

39. The method of claim 35, wherein a control population used in said classifier is an unbiased population of individuals having the illness.

40. The method of claim 35, wherein a control population used in said classifier is a population of individuals having the illness and not developing a critical and/or life threatening response to the illness.

41. The method of claim 35, wherein a control population in said classifier is a population of individuals who are normal.

42. The method of claim 35, wherein a control population in said classifier is a population of individuals wherein the majority of members of the control population do not have the illness.

43. The method of claim 38, wherein a control population in said classifier is an unbiased population of individuals having the illness.

44. The method of claim 38 wherein a control population in said classifier is a population of individuals having the illness and not developing a critical and/or life threatening response to the illness.

45. The method of claim 38, wherein a control population in said classifier is a population of individuals who are normal.

46. The method of claim 38, wherein a control population in said classifier is a population of individuals wherein the majority of members of the control population do not have the illness.

47. A composition comprising a collection of two or more antibodies and a suitable buffer, said composition capable of selectively binding to two or more proteins biomarkers from a sample isolated from a test individual , wherein the protein biomarkers are: complement fragment C5a (C5a), angiopoietin-1 (Ang-1), angiopoietin-2 (Ang-2), 10 kDa interferon gamma-induced protein (IP-10), soluble tyrosine kinase with immunoglobulin-like loop and epidermal growth factor domain-2 (sTie-2), soluble intercellular adhesion molecule-1 (sICAM-1), vascular endothelial growth factor A (VEGF), soluble vascular endothelial growth factor receptor 1 (sFlt-1), Chitinase-3-like protein I
(CH13L1), soluble triggering receptor expressed on myeloid cells-1 (sTREM-1), C-reactive protein (CRP), procalcitonin (PCT), angiopoietin-like 3 (Ang-like 3), complement factor D
(Factor D), or interleukin 18 binding protein (IL18bp) as set out in Table 1, and wherein the composition is used to (i)detect and quantify a level of each of said two or more protein biomarker in said sample from said test individual, wherein the test individual has not been diagnosed or differentially diagnosed with an illness which may be critical and/or life threatening, (ii) compare said quantified levels of said protein biomarkers in said test individual to control levels of said protein biomarkers from a control population (iii) determining the presence of differential levels for each biomarker in the comparison of step (ii) so as to make a determination that said individual is at a significantly increased risk of a critical and/or life threatening response to illness as compared with the control population.

48. The composition of claim 47, wherein said detecting and quantifying of step (i) utilizes one or more devices to transform the sample into data indicative of the levels of each of said two or more protein biomarkers which can be used to compare to the control population.

49. The composition of claim 47, wherein said one or more devices is an an enzyme linked immunoassay which is utilized so as to transform the sample into data.

50. The composition of claim 47, wherein said individual is subjected to a treatment protocol on the basis of the determination made in step (iii).

51. A method of determining the likelihood of an individual having a critical and/or life threatening response to an illness, wherein the test individual has not been diagnosed or differentially diagnosed with an illness which may be critical and/or life threatening prior to performing said method, said method comprising (i)detecting and quantifying a level of each of two or more protein biomarkers in a sample, wherein said protein biomarkers are: complement fragment C5a (C5a), angiopoietin-1 (Ang-1), angiopoietin-2 (Ang-2), 10 kDa interferon gamma-induced protein (IP-10), soluble tyrosine kinase with immunoglobulin-like loop and epidermal growth factor domain-2 (sTie-2), soluble intercellular adhesion molecule-1 (sICAM-1), vascular endothelial growth factor A
(VEGF), soluble vascular endothelial growth factor receptor 1 (sFlt-1), Chitinase-3-like protein 1 (CHI3L1), soluble triggering receptor expressed on myeloid cells-1 (sTREM-1), C-reactive protein (CRP), procalcitonin (PCT), angiopoietin-like 3 (Ang-like 3), complement factor D (Factor ID), or interleukin 18 binding protein (IL18bp) as set out in Table 1 (ii) comparing said quantified levels of said protein biomarkers to control levels of said protein biomarkers from a control population (iii) determining the presence of differential levels for each biomarker in the comparison of step (ii) so as to make a determination that said individual is at an increased risk of having a critical and/or life threatening response to said illness as compared with the control population.

52. The method of claim 51, wherein said detecting and quantifying of step (i) utilizes one or more devices to transform the sample into data indicative of the levels of each of said two or more protein biomarkers which can be used to compare to the control population.

53. The method of claim 52, wherein said one or more devices is an enzyme linked immunoassay which is utilized so as to transform the sample into data.

54. The method of claim 51, wherein said individual is subjected to a treatment protocol on the basis of the determination made in step (iii).

55. A method of determining the likelihood of an individual having a critical and/or life threatening response to an illness, wherein the test individual has not been diagnosed or differentially diagnosed with an illness which may be critical and/or life threatening prior to performing said method, said method comprising (i)detecting and quantifying a level of each of two or more protein biomarkers in a sample, wherein said protein biomarkers are: complement fragment C5a (C5a), angiopoietin-I (Ang-1), angiopoietin-2 (Ang-2), 10 kDa interferon gamma-induced protein (IP-10), soluble tyrosine kinase with immunoglobulin-like loop and epidermal growth factor domain-2 (sTie-2), soluble intercellular adhesion molecule-1 (sICAM-1), vascular endothelial growth factor A
(VEGF), soluble vascular endothelial growth factor receptor I (sFlt-1), Chitinase-3-like protein I (CHI3L1), soluble triggering receptor expressed on myeloid cells-1 (sTREM-1), C-reactive protein (CRP), procalcitonin (PCT), angiopoietin-like 3 (Ang-like 3), complement factor D (Factor D), or interleukin 18 binding protein (IL18bp) as set out in Table 1, (ii) utilizing the quantified levels of each said protein biomarker from said sample in a classifier (iii) making a determination as to whether said individual is anincreased risk of a life threatening response as compared with the control population.

56. The method of claim 55, wherein said detecting and quantifying of step (i) utilizes one or more devices to transform the sample into data indicative of the levels of each of said two or more protein biomarkers which can be used to compare to the control population.

57. The method of claim 56, wherein said one or more devices is an enzyme linked immunoassay which is utilized so as to transform the sample into data.

58. The composition of claim 55 wherein said individual is subjected to a treatment protocol on the basis of the determination made in step (iii).