Diagnostic markers for ischemia

CA 02425784 2003-04-15

WO 03/025571 PCT/US02/29477

DIAGNOSTIC MARKERS FOR ISCHEMIA

Background of the Invention

Field of the Invention

This invention relates to the diagnosis and monitoring of ischemia, including but not limited to myocardial and cerebral ischemia, by measuring the concentration of molecules that do not originate from the ischemic tissue but whose concentration in the blood and other fluids changes as a consequence of the ischemic state.

Description of the Related Art

Ischemia is a reduction in blood flow. This reduction may occur for a variety of reasons, including but not limited to thrombosis, embolism, aneurysm, spasm, or collapse of a blood vessel due to deterioration. Because of the reduction in blood flow the tissue that would otherwise be nourished may no longer receive sufficient nutrition to maintain cellular integrity, it also may not be able to remove sufficient amounts of cellular waste products and it may also result in inadequate exchange of blood gases such as oxygen and carbon dioxide. The inability to transfer sufficient oxygen to the cells (hypoxia) may have many of the same consequences of ischemia and can also be detected by the same methods claimed in this invention.

If ischemia persists for a sufficient time, that is, if oxygenated blood flow is not restored, then the cells of the tissue normally perfused by the blood flow, will begin to die. This may occur gradually over time and may be unnoticed until sufficient cell destruction has occurred so that the function of the organism is significantly impaired. An example may be the gradual deterioration of circulation to the extremities or other body parts that occurs in diabetes. The disruption of blood flow may also occur more acutely. This includes but is not limited to thrombus formation that results in reduction of blood flow through the coronary arteries, lodging of an embolism in a cerebral ischemia and similar events in the kidney and limbs.

For either a gradual or an acute time course, the earlier the ischemic/hypoxic condition can be detected and the sooner palliative therapy can be applied, the better will be the outcome for tissue and organism. For example, early detection of the diabetic-mediated ischemia in extremities might avoid amputation and early detection and relief of acute blockages to the heart or brain significantly reduces mortality and morbidity.

Unfortunately, early detection of ischemia/hypoxia is often not possible. For .example, the ECG which is the primary early diagnostic tool for acute coronary syndromes is less than 40% sensitive. For stroke patients the only tools available are either a CT scan or MRI, both of which can only determine if the patient has a hemorrhage in the first several hours after symptoms began. Only much later, when it is too late to administer the only treatment of ischemic stroke, thrombolytic therapy, are these imaging methods able to determine if an ischemic stroke has occurred. Thrombolytic therapy must be administered within 3 hours of

CA 02425784 2003-04-15

WO 03/025571 PCT/US02/29477 symptoms. For other organs there are no well established methods for early detection of ischemia. Thus a sensitive, accurate and rapid test for ischemia is needed for the diagnosis and treatment of patients.

In the final stages of ischemia, when cells begin to die (necrosis), they may release some of their contents into the blood. These are primarily intracellular proteins that are released because the normal barrier to containment, the cell’s membrane, is compromised by biochemical changes associated with death. Often these molecules are tissue specific, for example, cardiac troponin in the case of the heart. Although these molecules accurately reflect the presence of disease, they generally require several hours after symptoms occur to reach levels of significance in blood and they are only released from dead or dying cells. Thus in addition to sensitivity and accuracy an important feature of a test for ischemia would be its ability to detect the ischemic state well before necrosis.

In addition to the release of molecules from the ischemic tissue that are markers of necrosis, the ischemic event may generate a series of biochemical changes that can result in the change in concentration of molecules within the blood or other fluids that do not originate from the ischemic tissue. These are referred to in this invention as ischemic markers. The generation of ischemic markers can occur, for example, when the ischemic tissue generates molecules that are then converted to different molecules by a non-ischemic tissue or the ischemic tissue generates molecules that activate, from non-ischemic tissue, the release of different molecules into body fluids. Examples of this are the generation of norepinephrine, TNF_a and natriuretic peptides by ischemic cardiac or cerebral tissue. The norepinephrine, TNF_[X and natriuretic peptides so generated activate lipolysis in adipose tissue resulting in the elevation of free fatty acids in blood. In another example, sphingosine released from the ischemic myocardium is converted to sphingosine-1phosphate in platelets and may then be detected in blood (Yatomi, et al. (1997) Journ. Biol. Chem. vol. 272: pages 5291-5297; US Pat. No. 6,210,976).

In addition to diagnosing the presence of disease, levels of ischemic markers may predict risk of future deleterious events. Moreover, these molecules, at sufficient levels, may themselves mediate cellular effects that result In deleterious outcomes. For example, a large long term study of apparently healthy men revealed that increasing levels of total serum free fatty acids (FFA), although within the normal range, were associated with an increased risk of sudden death 22 years later. It was speculated that this increased rate of death was a consequence of FFA induced cardiac arrhythmias. In another example, use of the combination of glucose-insulin-potassium (GIK) in patients suffering from acute myocardial infarcts, produced a significant reduction in mortality relative to patients who did not receive GIK. One theory for this beneficial effect is the reduction in serum total FFA produced by GIK. Thus there is a need to be able to monitor ischemic markers to evaluate longer term risk of disease and to help to decide on the type of therapeutic intervention to reduce the increased risk associated with what may be a chronic ischemic state.

Summary of the Invention

CA 02425784 2003-04-15

WO 03/025571 PCT/US02/29477

The present invention describes the use in the diagnosis and monitoring of ischemia, of measuring the concentration in blood and/or other fluids of molecules that do not originate from the ischemic tissue. These molecules are called ischemic markers. The change in the concentration of these ischemic markers, relative to the non-ischemic condition, is indicative of some type of ischemia.

In one embodiment, the present invention is drawn to a method of detecting a condition which is indicative of ischemia in a mammal which includes the steps of:

(a) measuring a level of an ischemic marker in a test sample from a fluid of the mammal; and (b) determining if the level of the ischemic marker measured in the test sample correlates with ischemia in the mammal. In a preferred embodiment, the mammal is a human. The fluid used for the test sample may be selected from the group including but not limited to blood, serum, plasma, saliva, bile, gastric juices, cerebral spinal fluid, lymph, interstitial fluid or urine. In a preferred embodiment, the fluid used for the test sample is blood.

In a one embodiment, the ischemic marker is a lipid. The lipid may be selected from the group including, but not limited to, a sphingolipid, a lysolipid, a glycolipid, a steroid, and an eicosanoid, including leukotrienes, prostacyclins, prostaglandins and thromboxanes. In one embodiment, the sphingolipid used as the ischemic marker may be sphingosine, or a metabolite thereof including ceramide (Cer, Nacylsphingosine), sphingosine-1 -phosphate, sphingosylphosphorylcholine, or dihydrosphingosine, for example. In a more preferred embodiment, the lipid is a fatty acid.

The disclosed method may be used for detection of ischemia in any organ or tissue including but not limited to the heart, brain, kidney or a limb.

In one embodiment of the invention, a component of the ischemic marker which is soluble in an aqueous buffer is detected. In an alternate embodiment, the component of the ischemic marker which is not soluble in an aqueous buffer is detected.

The ischemic marker may be detected by spectroscopic means including but not limited to UV/VIS, infrared, microwave, radio, absorption or emission spectroscopes. Chromatographic procedures are also encompassed by the present invention, including but not limited to HPLC, low pressure chromatography, medium pressure chromatography, and gas chromatography. Detection of the ischemic marker by electron spin resonance using a spin label is also encompassed by the invention. In an alternate embodiment, the ischemic marker may be detected by an antibody or receptor molecule, immunoassay or enzymatic assay.

In a preferred embodiment, the method may include an initial step of selecting a mammal presenting symptoms of ischemia. In some embodiments, the method may also include the step of administering antiischemic therapy if the level of ischemic marker correlates with ischemia in said mammal. The anti-ischemic therapy may include a means to lower levels of serum fatty acids. In some embodiments, the anti-ischemic therapy is reperfusion therapy, antithrombolytic therapy, angiogenic therapy or surgery.

CA 02425784 2003-04-15

WO 03/025571 PCT/US02/29477

In some embodiments, the determination step is a comparison between said measured level of said ischemic marker and a predetermined value for the level of said marker. In some embodiments, the predetermined value for the level of the marker is indicative of the non-ischemic condition.

In one embodiment, the ischemic marker is an unbound or water-soluble free fatty acid. In a preferred embodiment, the unbound or water-soluble free fatty acid is detected by a protein that binds fatty acid. In a preferred embodiment, the protein that binds an unbound or water-soluble free fatty acid is fluorescent and exhibits a fluorescence that is different when the fatty acid is bound than when it is not bound. In some embodiments, the protein is one of the family of intracellular Fatty Acid Binding Proteins (FABPs) that have molecular weights between about 13,000 and 16,000 Dalton. In a preferred embodiment, the FABP is a rat intestinal FABP, in a preferred embodiment, the FABP is covalently labeled with acrylodan at position 27. In a more preferred embodiment, the acrylodan-labeled FABP is the leucine 72 to alanine mutant.

In some embodiments, a level of the unbound or water-soluble free fatty acid greater than 2 standard deviation units above an average value of a level of the unbound or water-soluble free fatty acid determined from a non-ischemic population is indicative of the ischemic condition. In some embodiments, a level of the unbound or water-soluble free fatty acid greater than about twice an average value of a level of the unbound or water-soluble free fatty acid determined from a non-ischemic population is indicative of the ischemic condition. In some embodiments, a level of the unbound or water-soluble free fatty acid greater than about 5 nM is indicative of the ischemic condition.

In an alternate embodiment, the protein that binds an unbound or water-soluble free fatty acid is albumin. In a preferred embodiment, the albumin is covalently labeled with 7-hydroxycoumarin or anthraniloyl.

In an alternate embodiment, the ischemic marker is total free fatty acid. In an alternate preferred embodiment a level of the total free fatty acid greater than 2 standard deviation units above an average value for the level of the total free fatty acid determined from a non-ischemic population is indicative of the ischemic condition. In an alternate preferred embodiment, a level of the total free fatty acid greater than about twice an average value of a level of the total free fatty acid determined from a non-ischemic population is indicative of the ischemic condition. In another embodiment, the ischemic marker is a ratio of total free fatty acid to albumin.

In some embodiments, the ratio of total free fatty acid to albumin greater than about two standard deviation units above an average value of the ratio of total free fatty acid to albumin determined from a nonischemic population is indicative of the ischemic condition. In some embodiments, the ratio of total free fatty acid to albumin greater than about twice an average value of the ratio of total free fatty acid to albumin determined from a non-ischemic population is indicative of the ischemic condition.

In one embodiment, a method of determining patient response to a treatment for ischemia is described which includes the steps of detecting ischemia as described above and determining if the level of

CA 02425784 2003-04-15

WO 03/025571 PCT/US02/29477 the ischemic marker is trending towards the level of the marker in the non-ischemic condition. In one embodiment, the treatment is reperfusion therapy which includes but is not limited to angioplasty or administration of a thrombolytic agent.

In one embodiment, a method of identifying patients at high risk for hemorrhage after receiving reperfusion therapy is described which includes the steps of:

measuring the level of a lipid component in a body fluid sample from the patient;

comparing the measured level of the lipid component from the patient to a threshold level of a lipid component, wherein the threshold level is determined from measuring a lipid component in body fluid of a normal population that does not have ischemia;

determining a ratio of the measured level of the lipid component from the patient to the threshold level; and correlating the ratio with the relative risk for hemorrhage after reperfusion therapy such that a high ratio indicates a high risk. In a preferred embodiment, the lipid component is selected from the group including unbound free fatty acid, total free fatty acid and a ratio of total free fatty acid to albumin.

In one embodiment, a method of identifying patients at high risk for mortality within three years after an ischemic event is described which includes the steps of:

measuring the level of a lipid component in a body fluid sample from the patient;

comparing the measured level of the lipid component from the patient to a threshold level of a lipid component, wherein the threshold level is determined from measuring a lipid component in body fluid of a normal population that does not have ischemia;

determining a ratio of the measured level of the lipid component from the patient to the threshold level; and correlating the ratio with the relative risk for mortality within three years after an ischemic event such that a high ratio indicates a high risk. In a preferred embodiment, the lipid component is selected from the group including unbound free fatty acid, total free fatty acid and a ratio of total free fatty acid to albumin.

In one embodiment, a method of identifying patients at high risk for hemorrhage is described which includes the steps of:

measuring a lipid component in a body fluid sample from the patient; .

comparing the measured level of the lipid component from the patient to a threshold level of a lipid component, wherein the threshold level is determined from measuring a lipid component in body fluid of a normal population that does not have ischemia;

determining a ratio of the measured level of the lipid component from the patient to the threshold level; and

CA 02425784 2003-04-15

WO 03/025571 PCT/US02/29477 correlating the ratio with the relative risk for hemorrhage such that a high ratio indicates a high risk. In a preferred embodiment, the lipid component is selected from the group including unbound free fatty acid, total free fatty acid and a ratio of total free fatty acid to albumin.

In another embodiment of the invention, a method is described for treating patients at high risk for hemorrhage after receiving reperfusion therapy including the steps of:

identifying a patient at high risk for hemorrhage after reperfusion therapy using the method described above; and treating said high-risk patient with an anti-ischemic therapy. In one embodiment, the anti-ischemic therapy is selected from the group including glucose-insulin-potassium compositions and albumin compositions.

In another embodiment a method of treating patients at high risk for mortality within three years after an ischemic event is described which includes the steps of:

identifying a patient at high risk for mortality within three years after an ischemic event using the method described above; and treating the high-risk patient with an anti-ischemic therapy. In one embodiment, the anti-ischemic therapy is selected from the group including glucose-insulin-potassium compositions and albumin compositions.

In another embodiment, a method of treating patients at high risk for hemorrhage is described which includes the steps of :

identifying a patient at high risk for hemorrhage using the method described above; and treating the high-risk patient with an anti-ischemic therapy. In one embodiment, the anti-ischemic therapy is selected from the group including glucose-insulin-potassium compositions and albumin compositions.

In one embodiment, a method of assessing long-term risk in non-acute patients is described which includes the steps of:

measuring the level of a lipid component in a body fluid sample from the patient;

comparing the measured level of the lipid component from the patient to a threshold level of a lipid component, wherein the threshold level is determined from measuring a lipid component in body fluid of a normal population that does not have ischemia;

determining a ratio of the measured level of the lipid component from the patient to the threshold level; and correlating the ratio with the relative risk for mortality such that a high ratio indicates a high risk, in a preferred embodiment, the lipid component is selected from the group including unbound free fatty acid and a ratio of total free fatty acid to albumin.

CA 02425784 2003-04-15

WO 03/025571

PCT/US02/29477

Further aspects, features and advantages of this invention will become apparent from the detailed description of the preferred embodiments which follow.

Brief Description of the Drawings

These and other feature of this invention will now be described with reference to the drawings of preferred embodiments which are intended to illustrate and not to limit the invention.

Figure 1 shows the level of unbound free fatty acid in blood samples taken from the Thrombolysis in Myocardial Infarction II (TIM! II) study of patients with AMI. Patients included in the study presented to the Emergency Department within 4 hours of onset of symptoms. Histogram representation shows the results of [FFAJ measurements in blood draws at times 0 (time of admission into Emergency Dept), 50 min., 5 hour and 8 hour for each of the first 250 patients, measured with ADIFAB2.

Figure 2 shows FFAu results in the TIMI II study of patients with AMI. Histogram representation shows the results for the second set of 250 patients measured with AD1FAB2, The results are virtually identical to those in Figure 1 for the first 250 patients.

Figure 3 shows creatine kinase (CK) levels in TIMI II patients at time 0 (admission time) (closed box, --) and at 4 hours (closed triangle, -A -). Values were determined by the TIMI investigators for patients corresponding to shipments 1 and 2. The upper limit of normal was determined for each day’s measurements and for each study site (open box,

Figure 4 shows mortality increases with [FFAJ values at admission in TIMI II patients. About 400 of the 500 patients in this set had blood draws at time of admission. The levels of FFAu were sorted and the number of deaths in each quartile were determined. The results yield a positive correlation that increases by more than 2 fold from lowest to highest quartile (p<0.042). Median [FFAJ for each quartile are shown in the bars.

Figure 5 shows the correlation between [FFAu] at admission and severe hemorrhagic events in TIMI II patients. The number of severe hemorrhagic events per quartile of FFA_U levels is shown. Numbers in the bars are the median [FFAJ in nM. The results indicate a more than 3 fold increase in risk from lowest to highest quartile with a p < 0.01.

Detailed Description of the Preferred Embodiment

While the described embodiment represents the preferred embodiment of the present invention, it is to be understood that modifications will occur to those skilled in the art without departing from the spirit of the invention. The scope of the invention is therefore to be determined solely by the appended claims.

The present invention describes the use, in the diagnosis and monitoring of ischemia, of measuring the concentration of molecules that do not originate from the ischemic tissue in blood and/or other fluids. These molecules are called ischemic markers. The change in the concentration of these ischemic markers,

CA 02425784 2003-04-15

WO 03/025571 PCT/US02/29477 relative to the non-ischemic condition, is indicative of some type of ischemia. A marker is considered not to originate from the ischemic tissue even though some fraction may have so originated, so long as that fraction represents only a small portion of the marker present in the sample, e.g., 20%, 10%, 5%, 2%, 1% or less.

In one embodiment, the ischemic marker is a lipid component such as a fatty acid. Total fatty acids may be determined by methods well known in the art such as taught in U.S. Patent No. 4,071,413, U.S. Patent No. 5,512,429, U.S. Patent No. 5,449,607, and U.S. Patent No. 4,369, 250 all of which are incorporated herein by reference. Levels of total fatty acid obtained by these methods may be compared with levels obtained from normal individuals in order to detect an ischemic condition in an individual. In a preferred embodiment, a ratio of total fatty acid to albumin is determined and compared with a normal population.

In another embodiment, unbound free fatty acid levels are measured in a bodily fluid as described below. Unbound free fatty acids (FFA_U) are the portion of free fatty acids soluble in the aqueous phase. In most body fluids free fatty acid (FFA) is mostly bound to proteins, for example, albumin, and membranes but a significant minority is unbound (FFA_U) and soluble in the aqueous phase. FFA_U are also referred to as watersoluble free fatty acids.

The bodily fluid may be selected from the group including cerebral spinal fluid, blood, serum, plasma, urine, saliva, lymph, gastric juices, interstitial fluid or bile In preferred embodiments, the fluid used for the test sample is taken from blood.

In one embodiment, any assay that provides an indication of the level of unbound free fatty acid (FFA_U) in body fluid relative to an asymptomatic population may be used in the diagnostic method to detect an ischemic condition. Preferably a threshold value is determined from a normal population that does not have an ischemic condition. In one embodiment, the threshold value is a concentration of FFAu in a body fluid that is significantly higher than the concentration of FFA_U in the body fluid of a control population that does not have an ischemic condition. In one embodiment, the threshold value is a concentration of FFA_U in a body fluid that is at least about two standard deviations greater than an average concentration of FFAu in a body fluid of a control population that does not have an ischemic condition. In another embodiment, the threshold concentration of FFAu in a body fluid is at least about 5 nM. In another embodiment, the threshold concentration of FFA_U in a body fluid is at least about twice the average concentration of FFAu in a body fluid of a control population that does not have an ischemic condition.

The present methods are particularly valuable when applied to selected patient populations. Thus, in one embodiment, the method is applied to a sample from an individual at risk for ischemia, such as diabetic patients, or surgery patients, or patients with familial or lifestyle risk factors. In another embodiment, the method is applied to a patient presenting symptomology consistent with ischemia or an ischemic event, or a patient suspected of having ischemia. Particular patient subpopulations that can be selected include patients having symptomology consistent with cardiac ischemia, brain ischemia, kidney ischemia, limb ischemia, or other clinically-significant ischemia. The method may be selectively applied to patients presenting

CA 02425784 2003-04-15

WO 03/025571 PCT/US02/29477 symptomology consistent with or inconsistent with one particular type of ischemia. Non-limiting examples would be selection of a patient having ischemic symptoms other than cardiac ischemia, or symptoms other than brain ischemia, or symptoms other than peripheral or diabetic-related ischemia. The present method may similarly be used to detect ischemia in cardiac patients in the absence of myocardial infarction by selecting an appropriate patient population. In one embodiment of any of the methods of the present invention, the patient is a mammal, and in a particular embodiment, the patient is a human.

In still a further embodiment of the invention, the method of the present invention further includes a treatment step. Thus, a patient is first tested for an ischemic condition, represented by elevation of one of the markers discussed herein. Patients having elevated levels of the marker are then treated with anti-ischemic therapy of any suitable type, such as reperfusion therapy, antithromboiytic therapy, angiogenic therapy, surgery, or the like. For example, patients may advantageously be tested within 10, 7, 5, or 3 hours of an ischemic event or symptom, and may then receive anti-ischemic therapy within 3, 2, or preferably 1 hour of the test.

In one embodiment, the present invention uses a fluorescently-labeled fatty acid binding protein (FABP) to measure an increased amount of unbound free fatty acid (FFA_U) in body fluid samples associated with an ischemic condition by quantitatively detecting a shift in fluorescence associated with binding of a FFA_Umolecule to the fluorescently-labeled FABP. In serum, free fatty acid (FFA) is mostly bound to albumin but a significant minority is unbound (FFAJ and soluble in the aqueous phase. This invention utilizes the method substantially as described in U.S. Patent No. 5,470,714, which is incorporated herein by reference. A variety of FABP and fluorescent labels can be used in detecting levels of FFA_U in body fluids such as blood. These include, but are not limited to, rat intestinal FABP (l-FABP), human adipocyte FABP (A-FABP) and rat heart FABP (H-FABP). Site-specific mutant forms of these FABP, in which one or more amino acid residues have been altered (inserted, deleted and/or substituted) are also useful in the method and include, for example, substitutions of Cys in l-FABP at positions 27, 81, 82, 84, an Ala substitution at residue 72 of l-FABP, and a Lys substitution at residue 27 of H-FABP. The FABP molecules may be fluorescently-labeled using a variety of known labels including but not limited to acryiodan, danzyl aziridine, 4-[N-[(2-iodoacetoxy)ethyl]-Nmethylamino]-7-nitrobenz-2-oxa-1,3-diazole ester (IANBDE), and 4-[N-[(2-iodoacetoxy)ethyl]-N-methylamino7-nitrobenz-2-oxa-1,3-diazole (IANBDA). Any label may be used in the practice of the invention as long as a measurable difference may be detected upon binding of a free fatty acid. For example, a difference in wavelength, signal intensity, or polarization, or lifetime may be monitored. Further examples of labels which may be used include, but are not limited to, chromophores which produce a change in absorbance or optical activity and spin labels which change is detectable by electron spin resonance. In a preferred embodiment, a fluorescently-labeled FABP is acrylodan-derived recombinant rat intestinal fatty acid binding protein (referred to as ADIFAB). Derivatization with acryiodan was performed using known methods substantially as previously described (U.S. Patent No. 5,470,714 & Richieri, G.V, et al., J. Biol. Chem., (1992) 276: 234959

CA 02425784 2003-04-15

WO 03/025571 PCT/US02/29477

23501), and ADIFAB is commercially available (FFA Sciences LLC, San Diego, CA ). Concentrations of FFA_U can be determined by the binding of the FFAu to the fluorescently labeled fatty acid binding protein (FABP). A different fluorescence is exhibited when no FFA is bound to the fluorescently-labeled FABP. The concentration of FFA_U can be determined from the fluorescence difference. The wavelength emitted by the 5 fluorescently-labeled FABP depends upon the label and protein used. In one embodiment, the protein is either l-FABP or a l-FABP with a substitution of Ala for Leu at position 72 where the label is acryolodan. These species are referred to as ADIFAB and ADIFAB2, respectively.

Briefly, ADIFAB2 was obtained as follows. First, a restriction fragment, carrying the Ala⁷² mutation and appropriate complementary ends, was substituted for the wild-type restriction fragment spanning the Sall 10 site at position 211 in the l-FABP cDNA sequence (Alpers, et al. (1984) Proc. Natl. Acad. Sci U.S.A. 81:313317) and a Pme\ site at position 291, which was introduced by site-specific mutagenesis. The mutated restriction fragment was constructed by annealing partially complementary synthetic oligonucleotides carrying the desired sequence, filling in single-stranded regions with the DNA polymerase Klenow fragment, and digesting with Sa/I and Pme\ to give the appropriate termini. The Ala⁷² substituted l-FABP cDNA was inserted 15 into the pET11 vector and was expressed in the BL21(DE3) strain. The mutant l-FABP was purified essentially by the method of Lowe, et al. (Lowe, et al. (1987) J. Biol. Chem. 262: pages 5931-5937) and yielded about 100 mg of purified protein per liter of Escherichia coli culture. Acrylodan derivatization was done as described previously for ADIFAB (see U.S. Patent No. 5,470,714, incorporated herein by reference). Lipidex-5000 chromatography was used to remove free acrylodan.

The binding affinities of ADIFAB2 have been found to be about 10-fold greater than ADIFAB. The wavelength emitted by these fluorescently-labeled FABP's is about 420 to 470 nm. The emission wavelength for the FFA bound to the fluorescently-labeled FABP is about 495 to 580 nm. It will be understood that those skilled in the art can readily substitute other fluorescently-labeled FABP in the disclosed assay.

The assay for determination of FFA_U levels in aqueous samples measures the intensity of a shift in 25 fluorescence from a first wavelength, at which the, derivatized FABP fluoresces when no FFA is bound, to a second wavelength, at which the derivatized FABP fluoresces when a molecule of FFA is bound, and the concentration of FFA_U is then determined from the ratio (R value) of the two intensities of fluorescence wavelengths as described in U.S. Patent No. 5,470,714 and Richieri, et al, (1995) J. Lipid Research, vol. 36: pages 229-240, both of which are incorporated herein by reference. Briefly, the ratio is calculated using the 30 following formula:

1(1)- l(1)blank

R= __________________________

1(2)-l(2)b!ank

CA 02425784 2003-04-15

WO 03/025571 PCT/US02/29477 wherein, 1(1) and l(1)blank are the measured fluorescence intensities for a sample containing ADIFAB or ADIFAB2 and a blank sample containing all reagents except ADIFAB or AD1FAB2, respectively at wavelength 1; and [(2) and l(2)blank are the corresponding fluorescence intensities at wavelength 2. For ADIFAB, wavelength Ί is in preferably (but not absolutely) in the range of 495 and 515 nm and wavelength 2 is in the range from 422 to 442 nm. Measurements of fluorescence intensities may be obtained using standard techniques. As recognized by those skilled in the art, under appropriate conditions, the Iblank intensities in the above equation can be neglected.

Quantitative detection of levels of body fluid FFA_U that are elevated over body fluid FFA_U levels found in normal healthy individuals can be used to diagnose an ischemic condition, in one embodiment, an isGhemic condition is indicated by a concentration of unbound or water-soluble free fatty acid in a body fluid sample that is significantly higher than the concentration of unbound or water-soluble free fatty acid in a body fluid sample of a control population that does not have an ischemic condition. In one embodiment, an ischemic condition is detected by levels of unbound or water-soluble free fatty acid in a body fluid sample that exceed the average normal levels of unbound or water-soluble free fatty acid in a body fluid sample by about 2 standard deviations. In another embodiment, unbound or water-soluble free fatty acid levels which are greater than about 5 nM are indicative of an ischemic condition. In another embodiment, a level of unbound or water-soluble free fatty acid in a body fluid sample greater than about twice the average value of unbound or water-soluble free fatty acid determined from a body fluid sample from a non-ischemic population is indicative of the ischemic condition. Elevated levels of unbound or water-soluble free fatty acid in a body fluid sample which are considerably higher may also be detected.

The diagnostic method of the present invention is based on the discovery that patients experiencing an ischemic condition have elevated levels of lipids such as FFA_U in body fluids such as blood, compared to normal levels of body fluid FFAu in healthy individuals. While the diagnostic method may be carried out at any time, preferably, the test is carried out within 24 hours of the ischemic event. More preferably, the test is carried out within 10 hours of the ischemic event. Most preferably, the test is carried out within 3 hours of the ischemic event so that treatment may be initiated. Many treatments for ischemic conditions such as cardiac ischemia must be initiated quickly (within hours) in order to be effective. Treatment without a correct diagnosis can be most deleterious to the patient. The ability to quickly diagnose an ischemic condition is an advantage of the invention described here.

Briefly, the FFA_U assay and determinations were performed as follows. Blood samples were diluted 100-fold in buffer (20 mM N-2-hydroxyethyl-piperazine-N-2-ethane sulfonic acid (HEPES), 150 mM NaCI, 5 mM KCl and 1 mM I^HPCX adjusted to pH 7.4), yielding a serum albumin concentration of about 6 μΜ. A solution of fatty acid-free albumin plus ADIFAB or ADIFAB2 was the negative control. For each donor, two aliquots of serum were prepared: one background or blank sample of 1% serum and one experimental

CA 02425784 2003-04-15

WO 03/025571 PCT/US02/29477 sample of 1% serum plus ADIFAB or ADIFAB2. The negative control, blank and experimental samples were measured at 22°C or 37°C.

For each sample, multiple measurements of pairs of intensities were collected at about 432 nm and 505 nm, for ADIFAB and about 450 and 550 for ADIFAB2 and R values were determined after subtraction of the intensities of the blank sample. At least two separate measurements were done on different days for each serum sample and the mean values and standard deviations of FFA_U concentrations were determined. To determine the probabilities of a difference between sets of measures, differences in means were evaluated using Student's t test, where a p value of less than 0.05 was considered significant.

The present disclosed method for detecting ischemia may further comprise the steps of taking measures to reduce the levels of ischemic markers. Also encompassed by the present disclosure is a method of determining prognosis of a patient following treatment for ischemia by detecting ischemia as disclosed herein and determining if the level of the ischemic marker is trending towards the level of the marker in the non-ischemic condition. The disclosed method also comprises the evaluation of the risks associated with elevated ischemic markers, either because the elevation of these markers reflects an underlying pathologic state and/or because the ischemic markers are themselves deleterious.

Example 1: [FFA_U1 are elevated before troponin I in patients presenting to the Emergency department (ED) with different forms of myocardial ischemia

To determine the concentration of unbound free fatty acids [FFAJ in patients suffering from a wide range of natural” ischemic insults we measured FFA_U levels in patients presenting to the emergency department (ED) of the Hennepin County Medical Center (Minneapolis, MN). Samples were drawn at admission and several times later from 29 patients. Each of these patients was diagnosed as having cardiac ischemia, primarily by ST segment changes. The diagnoses for these patients included: acute myocardial infarction (AMI), congestive heart failure (CHF), unstable angina, cardiac contusion, cardiac surgery, hypertension and cocaine induced ischemia. FFA_U were significantly elevated in every patient diagnosed with cardiac ischemia, on average 20 times higher than normals (this study also included 48 “normal·¹ patients whose average FFA_U levels were 3 nM).

FFAu levels were elevated at the earliest draw (admission to the ED) in every patient, including the 9 AMI patients in this study (Table). In contrast, 7 of these 9 AMI patients showed no elevation of the cardiac marker Troponin I (Tnl) at the time of admission. In these 7 patients Tnl became elevated between 4 and 32 hours after admission. Furthermore, most of the non-AMI patients did not reveal elevated Tnl levels, consistent with [FFA_U] being a marker of ischemia before necrosis.

Another noteworthy feature of these results is that in the absence of reperfusion therapy, FFA_U levels remain elevated for long periods (up to days) after the initial ischemic event (Table). Thus FFA_U respond rapidly to an ischemic event but return only slowly (presumably as the ischemia resolves) to base line, thereby providing a uniquely wide diagnostic window.

CA 02425784 2003-04-15

WO 03/025571

PCT/US02/29477

TABLE: Time course of Tnl and [FFA_U] in AMT

<td> Patient</td><td> TIM E</td><td> Tnl</td><td> [FFAu]</td>

<td></td><td> (hour s)</td><td> (ng/ml)</td><td> (nM)</td>

<td rowspan="2"> 1</td><td> 0</td><td> 0.3</td><td> 314</td>

<td> 4</td><td> 0.8</td><td> 27</td>

<td rowspan="3"> 2</td><td> 0</td><td> 0.3</td><td> 23</td>

<td> 54</td><td> 7.1</td><td> 463</td>

<td> 62</td><td> 15.9</td><td> 797</td>

<td rowspan="4"> 3</td><td> 0</td><td> 0.3</td><td> 22</td>

<td> 5</td><td> 0.3</td><td> 282</td>

<td> 29</td><td> 5.2</td><td> 14</td>

<td> 36</td><td> 4.5</td><td> 19</td>

<td rowspan="4"> 4</td><td> 0</td><td> 0.3</td><td> 12</td>

<td> 5</td><td> 0.3</td><td> 19</td>

<td> 9</td><td> 0.3</td><td> 9</td>

<td></td><td> 2.7</td><td> 14</td>

<td rowspan="3"> 5</td><td> 0</td><td> 0.3</td><td> 22</td>

<td> 16</td><td> 2.1</td><td> 29</td>

<td> 22</td><td> 2.5</td><td> 54</td>

<td rowspan="2"> 6</td><td> 0</td><td></td><td> 9</td>

<td> 7</td><td> 0.6</td><td> 22</td>

<td rowspan="2"> 7</td><td> 0</td><td> 2.</td><td> 8</td>

<td> 10</td><td> 3.1</td><td> 112</td>

<td rowspan="2"> S</td><td> 0</td><td> ' 0.4</td><td> 18</td>

<td> 13</td><td> 1</td><td> 19</td>

<td rowspan="6"> 9</td><td> 0</td><td> 10.6</td><td> 7</td>

<td> 7</td><td> 6.3</td><td> 11</td>

<td> 21</td><td> 6.9</td><td> 36</td>

<td> 27</td><td> 5.4</td><td> 25</td>

<td> 38</td><td> 4.2</td><td> 26</td>

<td> 50</td><td> 2.1</td><td> 10</td>

^a: Hennepin AMI patients. Time is hours from time of admission. Baseline Tnl values were 0.3 ng/ml and normal [FFA_U] were 3 nM. Two or more serial measurements were done for each patient (separated by thick lines).

CA 02425784 2003-04-15

WO 03/025571 PCT/US02/29477

These results indicate the potential power of the FFA_U assay; a serious ischemic event can be detected hours to days before current cardiac markers indicate disease, raising the possibility of earlier intervention and significant improvement in patient outcome.

Example 2. Results from the Thrombolysis in Myocardial Infarction II (TIMI-II) trial indicate high sensitivity and specificity for detection of AMI using ÎFFAJ measurements.

To obtain more detailed information about the FFA_U response in acute myocardial infarction, we used blood specimens from the TIMIII trial (TIMI Study Group, New England Journal of Medicine (1989) vol. 320, page 618-625). These specimens have been maintained by the National Heart Lung Blood Institute and were made available for our investigations of FFAu levels in ischemia. Patients were enrolled in the TIMI ll trial if they presented to an ED with chest pain within 4 hours of the initial symptoms and exhibited an ST segment elevation. Every patient was treated intravenously over 6 hours with either 100 mg (90.5% of the 3262 patients treated) or 150 mg (9.5% of the 3262 patients treated) t-PA starting within about 10 minutes of enrollment. Blood samples were drawn upon admission (just before t-PA), 50 minutes after t-PA, and then 5 h and 8 h after start of t-PA treatment. A set of 4 such samples was stored at -70°C for each of approximately 2,500 patients. For each patient, the TIMI investigators recorded about 825 clinical, physical and chemical observations.

Results for the first 500 patients provide important insights into the relationship between acute cardiac ischemia and FFA_U levels. [Samples were received in individual shipments of 1000 samples (250 patients). Figures 1 and 2 show results for shipments 1 and 2, respectively.] (Kleinfeld, et al. (2002) J. Am. Coll. Cardiol. 39:312A).

Upon admission the average FFA_U (AD1FAB2) value was 13 nM, more than 14 standard deviations above the norma! value (2.8 + 0.7 nM). This result demonstrates clearly that [FFA_U] are highly elevated, early in the ischemic event. Using a 5 nM cutoff, the predicted sensitivity for detection of AMI was 91 % using FFA_Uat admission (using FFA_U at admission and at 50 minutes increases sensitivity to &gt; 98%). The results show that &gt; 98% of the patients have FFA_U levels &gt; 3 standard deviation above normal levels when the first two time points are included (&gt;91% with only 1^st time point).

At the same time, CK values at time 0 are generally less than the upper limit of normal (Figure 3). Creatine kinase levels at admission were elevated in less than 19 % of the patients and even in these patients the increases were modest, on average &lt; 50% above the upper limit of normal for the test. Not until 4 h post admission were creatine kinase levels elevated substantially in a majority of the patients.

At 50 minutes the average FFA_U (22 nM) is substantially higher than the already elevated initial values. These samples were obtained about 50 minutes after administration of t-PA (Figures 1 and 2). The significantly higher average FFAu levels may reflect reperfusion effects and the ischemia mediated increases 14

CA 02425784 2003-04-15

WO 03/025571 PCT/US02/29477 in [FFA_U] with increasing time. In any event, this time course appears to be changed abruptly by t-PA treatment because FFAu levels at 5 and 8 hours reveal significant decreases following reperfusion; the average value at 8 hours is significantly (p &lt; 0.05) lower than the FFA_U level at time of admission. This biphasic response appears to be correlated with reperfusion because the AMI patients in the Hennepin study who did not receive reperfusion therapy reveal [FFA_U] that remained elevated for times greater than 24 h (data not shown). These findings, that FFAu may increase in response to reperfusion and then decrease within 5 h, indicate that FFA_U levels might be used to evaluate the success of reperfusion therapy.

The [FFAu] response of these patients appears to be an intrinsic and independent property of the pathology. Figures 1 and 2 reveal that measurements of the [FFAu] distributions at each of the 4 different sample times, yield virtually identical parameters for the distributions of 2 different sets of 250 patients; samples in shipments 1 and 2 are for two sets of different patients. Moreover, almost all patients with confirmed Mis revealed FFA_U (50 minutes) &gt; [FFA_U] (admission), whereas for most of the approximately 3.3% of patients without confirmed Ml, [FFAJ (50 minutes) &lt; [FFAJ (admission).

Example 3. High levels of [FFA_U1 in TIMl-ll trial samples at time of admission correlate with higher mortality 3 years post admission.

[FFA_U] values at the time of admission are correlated with a more than 2-fold increase in mortality. [FFAJ values for specimens drawn at time of admission (about 400 of the 500 patients) were sorted and partitioned into quartiles. The number of deaths, by any cause up to about 3 years after admission, were counted for each quartile. The primary (75%) cause of death was cardiovascular disease. The results show a more than 2 fold increase in death rate from lowest to highest quartile with a p value &lt; 0.025 (Figure 4). In contrast, the cardiac marker, creatine kinase, was normal, at the time of admission for most (&gt;80%) of these patients (Figure 3). Furthermore, [FFAJ values were not correlated with age, gender, race, weight, other disease, medications or systolic blood pressure on admission. These results indicate that in addition to their other diagnostic characteristics, FFAu levels measured at time of admission to the ED can be used to stratify patients according to their mortality risk (Kleinfeld, et al. (2002) J. Am. Coll. Cardiol. 39:312A).

Example 4. High levels of [FFAu] in TIMl-ll trial samples at time of admission correlate with increased rate of severe hemorrhage after t-PA therapy.

[FFAu] at admission also predict increased rates of severe hemorrhagic events after t-PA therapy. A similar analysis as for mortality (Example 3) was performed for severe hemorrhagic events. The number of patients that experienced severe hemorrhagic events was determined for each quartile (Figure 5). The results reveal a more than 3 fold increase in the rate of severe hemorrhagic events from lowest to highest quartile of the distribution of [FFAJ values at admission (before the start of t-PA administration) with p &lt; 0.01. One of the possibilities raised by these results is that rates of hemorrhage mediated by t-PA might be reduced if FFAu levels were lowered.

CA 02425784 2003-04-15

WO 03/025571 PCT/US02/29477

Although the results of Figures 4 and 5 do not indicate whether elevated FFA_U levels directly and independently contribute to increased rates of death and hemorrhage, abundant in vitro work demonstrates that elevated FFA are potent perturbera of many cellular functions. In addition, clinical evidence for a direct role of FFA is strongly suggested by many studies pointing to their arrhythmogenic role and their potential role in cardiovascular disease. Results from our TIM!II trial also suggest that [FFAJ are a strong independent factor of adverse outcomes. Thus, [FFAJ do not simply reflect the demographics of this patient population; no correlation was observed with age, gender, race, or weight, for example. Moreover, the correlations of [FFAJ at admission with mortality and hemorrhagic events, appear to be independent; although the mortality rate was higher for patients who experienced severe hemorrhagic events (22% vs 11% for all patients), only 15% of deaths were due to hemorrhage.

Example 5. Blood levels of FFA in ischemic patients do not originate from the ischemic tissue.

Cerebral and myocardial ischemia in humans result in increased plasma FFA levels by more than 7 fold above normal (Kleinfeld, et al. (1996) Amer. J. Cardiol. 78:1350-1354; Kurien, et al. (1966) The Lancet 16:122-127; Oliver, et al. (1994) The Lancet 343:155-158). It is also well known that FFA accumulate within the cells of an ischemic tissue, at least for isolated organs (Bazan, et al. (1970) Biochim. Biophys Acta 218: ΙΙΟ; Van derVusse, et al. (1997) Prostaglandins, Leukotrienes and Essential Fatty Acids 57:85-93). However, this generation of FFA by ischemic tissue cannot account for the increases in plasma FFA observed in ischemia. This follows because the amount of FFA generated by the ischemic tissue, even under conditions of complete ischemia of the whole organ, is negligible compared to normal plasma FFA levels, let alone the more than 7 fold increases observed in ischemic patients. For example, under normal conditions plasma FFA turns over every 2 min, which at (normal) plasma total [FFA] of about 500 μΜ translates into about 0.2 g of FFA/min. The FFA produced in ischemic tissue derives from phospholipid (Bazan, et ai. (1970) Biochim. Biophys Acta 218:1-10; Goto, et al. (1988) Stroke 19: 728-735; Jones et al. (1989) Am. J. Pathol. 135: 541556), which comprises less than 1 % by (wet) weight of a typical (non-adipose) cell. Therefore, under ischemic conditions, where total FFA is 2 to 7 fold greater than the normal level, more than 20 g/min. of tissue would have to be used to maintain the observed levels of FFA. This is not compatible with life and indicates that increases in plasma FFA are derived from non-ischemic tissue.

In the specific example of complete cerebral ischemia (decapitation) the amount of total FFA that accumulates in rat brain over a period of 30 minutes is about 0.8 pmoles (Ikeda, et al. (1986) J. Neurochem. 47:123-132). The total amount of plasma FFA in a 300 g rat is greater than 5 pmoles (&gt;10 ml plasma of 500 μΜ FFA). In a rat model of reversible middle cerebral artery ischemia (a much less profound ischemic insult than decapitation), total plasma FFA increases about 2 fold and therefore an additional 5 μϊηοΐθ3 of FFA is added to plasma every 2 minutes or about 75 μηιοίβε in 30 minutes. Thus the 0.8 pmoles produced in 30 minutes of complete ischemia is negligible in comparison to the amount of FFA needed to raise plasma levels to those observed in ischemia. The most likely source of tissue with sufficient capacity to generate the large

CA 02425784 2003-04-15

WO 03/025571 PCT/US02/29477 quantities of FFA observed in cerebral ischemia is adipose tissue. We have found that within minutes of cerebral ischemia increases are observed in plasma levels of TNF_a, a potent activator of adipose lipolysis (Ryden, et al. (2002) J. Biol. Chem. 277:1085-1091) and that this TNF_a is well correlated with FFAu increases in plasma.

It will be understood by those of skill in the art that numerous and various modifications can be made without departing from the spirit of the present invention. Therefore, it should be clearly understood that the forms of the present invention are illustrative only and are not intended to limit the scope of the present invention.