Methods for detecting and treating cavernous angioma

Abstract

The disclosure provides methods for diagnosing, prognosing, and treating cavernous angioma (CA). Also provided are methods for determining the efficacy of therapeutic treatments for CA, such as those that may be in a clinical trial setting or administered by a health care professional. Aspects of the disclosure relate to a method for treating cavernous angioma (CA) in a subject, the method comprising administering a therapy to a subject that has been evaluated for one or more metabolites and / or one or more biomarkers in a biological sample from the subject. Aspects relate to a method for prognosing or diagnosing CA in a subject, the method comprising evaluating one or more metabolites and / or one or more biomarkers in a biological sample from the subject. Aspects relate to a method for evaluating a subject comprising evaluating one or more metabolites and / or one or more biomarkers in a biological sample from the subject.

Description

METHODS FOR DETECTING AND TREATING CAVERNOUS ANGIOMA

[0001] This application claims priority of U. S. Provisional Application No. 63 / 711,859 filed October 25, 2024, which is hereby incorporated by reference in its entirety.DESCRIPTION

[0002] This invention was made with government support under grant number NS 114552 awarded by the National Institutes of Health. The government has certain rights in this invention.I. Field of the Invention

[0003] The present invention relates generally to the fields of molecular biology, neurology, and therapeutic diagnosis.II. Background

[0004] Cerebral cavernous malformations, also known as cavernous angiomas (CA) affect up to a million Americans, but fewer than 200,000 have suffered a symptomatic hemorrhage (SH). Yet those cases with SH are 10X more likely to rebleed and cause serious disability. And it is not always clear if a lesion has recently bled despite imaging and clinical symptoms, and further unclear which lesion would bleed in the near future. Hence it is a critical clinical need to identify CA cases that have recently bled and those which will imminently bleed. These would be targeted to closer surveillance and potential medical or surgical treatments. Conversely, those cases with CAs that have not bled nor a prone to bleeding can be reassured and watched safely. Circulating biomarkers are being sought to enhance the diagnosis and prognosis of hemorrhagic CAs, including proteins and microRNAs.SUMMARY OF THE INVENTION

[0005] The current disclosure provides for methods for diagnosing, prognosing, monitoring, and / or treating cavernous angioma (CA), also refers as Cerebral Cavernous Malformation (CCM) or Cavernous Malformation (CM). The current disclosure also provides methods for diagnosing and prognosing a patient as having, or not having, cavernous angioma with symptomatic hemorrhage (CASH). Also provided are methods for determining the efficacy of therapeutic treatments for CASH, such as those that may be in a clinical trial setting or administered by a health care professional.

[0006] Accordingly, disclosed herein are methods for treating CA and / or CASH in a patient; methods for measuring biomarkers in a patient having, suspected of having, or diagnosed with having CA and / or CASH; methods for prognosing a patient with CA and / or CASH; methods for diagnosing a patient with CA and / or CASH; methods for monitoring patients having,300224353.1 - 1 -suspected of having, or diagnosed with having CA and / or CASH; methods for prognosing, diagnosing, or monitoring CA, aggressive CA, non-aggressive CA, CASH, non-CASH, familial CA, or sporadic CA in a patient; methods of distinguishing a patient as having CASH or non-CASH; and methods of monitoring a patient with CA and / or CASH, including before, during, or after a treatment is provided to the patient. In some aspects, non-CASH is a cavernous angioma that does not include, or is not identified as including, symptomatic hemorrhage. In some aspects, the method comprises one or more steps including: measuring an amount of one or more biomarkers in a biological sample obtained from a patient; comparing the amount of a measured biomarker to a control; training a machine learning model on a training set of measured biomarkers; applying a trained model on the amount of measured biomarkers in the biological sample; determining the severity of CA and / or CASH in the patient; determining the likelihood the patient will develop CA and / or CASH; determining a treatment response of the patient; administering a therapy to a patient; and / or altering a treatment regime of a patient. The biomarkers may be metabolites and / or proteins, including any of the metabolites and / or proteins disclosed herein.

[0007] Further aspects relate to methods for evaluating a subject comprising evaluating one or more biomarkers in a biological sample from the subject. The biomarkers may be metabolites and / or proteins, including any of the metabolites and / or proteins disclosed herein.

[0008] Also described are kits comprising 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, or more detection agents (or any range derivable therein) for determining levels of one or more biomarkers for CA, including any of the biomarkers disclosed herein.

[0009] In some aspects, the biomarkers comprise or consist of tyrosine, acetyl-L-camitine, arachidonic acid, hexanoylcarnitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, vascular endothelial growth factor- A (VEGF), interleukin (IL)- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, thrombospondin- 1 (TSP1), thrombospondin-2 (TSP2), thrombomodulin (TM), cluster of differentiation- 14 (CD 14), c-reactive protein, roundabout guidance receptor 4 (ROBO4), lipopolysaccharide binding protein (LBP), or a combination thereof. In some aspects, the biomarkers comprise or consist of ROBO4, CD14, TM, and acetyl-L-carnitine. In some aspects, the biomarkers comprise or consist of ROBO4, CD14, and TM. In some aspects, the biomarkers comprise or consist of endoglin, pipecolic acid, arachidonic acid, and hypoxanthine. In some aspects, the biomarkers comprise or consist of endoglin, TSP2, and IL- 16.300224353.1 - 2 -

[0010] In some aspects, one or more of tyrosine, acetyl-L-camitine, arachidonic acid, hexanoylcarnitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, vascular endothelial growth factor-A (VEGF), interleukin (IL)- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, thrombospondin- 1 (TSP1), thrombospondin-2 (TSP2), thrombomodulin (TM), cluster of differentiation- 14 (CD 14), c-reactive protein, roundabout guidance receptor 4 (ROBO4), and lipopolysaccharide binding protein (LBP) are specifically excluded from a method, kit, or system described herein. In some aspects, one or more of tyrosine, acetyl-L-camitine, arachidonic acid, hexanoylcarnitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, vascular endothelial growth factor-A (VEGF), interleukin (IL)- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, thrombospondin- 1 (TSP1), thrombospondin-2 (TSP2), thrombomodulin (TM), cluster of differentiation- 14 (CD 14), c-reactive protein, roundabout guidance receptor 4 (ROBO4), and lipopolysaccharide binding protein (LBP) are not measured in a method described herein.

[0011] The biomarker may or may not be determined to have differential levels relative to a control. The biomarker may or may not be determined to have or evaluated as having levels not significantly different than a control. The biomarker may or may not be determined to be or evaluated as being increased relative to a control. The biomarker may or may not be determined to be or evaluated as being decreased relative to a control. The biomarker may or may not be evaluated or determined as having differential levels relative to a control. The biomarker may or may not be evaluated as having levels not significantly different than a control.

[0012] The subject may be one that has been diagnosed with CA. The subject may be one that has not been diagnosed with CA. The CA may be further defined as CA with symptomatic hemorrhage (CASH). The CA is further defined as CA without symptomatic hemorrhage (non-CASH).. CA may be further defined as familial CA or sporadic CA. Familial CA may be further define as familial-CCMl, familial-CCM2, familial-CCM3 or familial -unknown genotype. Familial CA may be further defined as familial-CCM3. The subject may be one that is being treated with a therapy for CA. The subject may be one is not being treated with a therapy for CA.

[0013] The sample from the subject may comprise a tissue sample, a blood sample, a whole blood sample, a fractionated sample, a plasma sample. The sample from the subject may comprise a plasma sample from the subject. The methods may exclude detection or300224353.1 - 3 -determination of any other biomarkers in the sample. The control may comprise the level of the metabolite or biomarker in samples from subjects identified as non-CASH, low risk, or as not having CA. The control may comprise the expression or protein level of the biomarker in samples from subjects identified as CASH, high risk, or as having CA. The control may comprise the level of the metabolite or biomarker in samples from subjects identified as having non-CASH, in subject identified as low risk, in subjects identified as not having CA, subject identified, or in subjects having familial CA. The control may comprise the expression or protein level of the biomarker in samples from subjects identified as high risk, in subject identified as having CASH, in subjects identified as having CA, in subject identified as, or in subjects identified as having sporadic CA. The treatment may comprise a gene therapy, a B-cell immunomodulation therapy, a statin therapy, a ROCK modulation therapy, surgical excision of a CA lesion, a VEGF inhibitor, or combinations thereof.

[0014] The subject may be diagnosed as having CA, diagnosed as not having CA, diagnosed as having CASH, diagnosed as non-CASH, prognosed as high risk for CASH, prognosed as low risk for CASH, diagnosed as having familial CA, or diagnosed as having sporadic CA when the measured biomarkers are altered relative to a reference or control value.

[0015] The difference in the level of a biomarker from a control may be, for example, a statistically significant difference. In some aspects, the difference is at least 0.5, 1, 1.5, 2, 2.5, or 3 (or any range or value derivable therein) standard deviations different than a control. The level may be a difference that is not significantly or statistically different than a control or that is within 2, 1.5, 1, or 0.5 (or any range or value derivable therein) standard deviations from a control. The biomarker may be determined or measured as increased relative to a control. The increase may be by at least, at most, approximately, or exactly 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 120, 140, 160, 180, 200, 250, 300, 350, or 400% (or any range or value derivable therein) or by at least, at most, approximately, or exactly 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 1920, 25, 30, 40, 50, 60, 70, 80, 90 100 fold (or any range or value derivable therein). The biomarker may be determined or measured as decreased relative to a control. The decrease may be by at least, at most, approximately, or exactly 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 120, 140, 160, 180, 200, 250, 300, 350, or 400% (or any range or value derivable therein) or by at least, at most, approximately, or exactly 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 1920, 25, 30, 40, 50, 60, 70, 80, 90 100 fold (or any range or value derivable therein).

[0016] In some aspects, the control comprises the level of the biomarker in samples from subjects identified as low risk. In some aspects, the control comprises the level of the biomarker300224353.1 - 4 -in samples from subjects identified as non-CASH. In some aspects, the control comprises the level of the biomarker in samples from subjects identified as CASH. In some aspects, the control comprises the level of the biomarker in samples from subjects identified as high risk. In some aspects, the control comprises the level of the biomarker in samples from subjects identified as not having CA. In some aspects, the control comprises the level of the biomarker in samples from subjects identified as having CA. In some aspects, the control comprises the level of the biomarker in samples from subjects with cases without CASH, low risk, or without CA. In some aspects, the control comprises the level of the biomarker in samples from subjects identified as having a CA that would bleed or grow (e.g. high risk) in the subsequent year after the sample is collected compared to a subject (or subjects) that does not grow or bleed in the subsequent year (e.g. low risk).

[0017] In some aspects, the control comprises the level of the biomarker measured in a different biological sample obtained from the patient. The different biological sample may be taken at least, at most, approximately, or exactly 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, or more (or any range or value derivable therein) hours, days, weeks, months, or years before the biological sample. In some aspects, the biological samples are obtained every at least, at most, approximately, or exactly 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, or more (or any range or value derivable therein) hours, days, weeks, months, or years. In some aspects, biological samples are obtained from the patient during a treatment regime (which may be a clinical trial treatment regime) and the biomarker(s) are measured in each biological sample obtained from the patient. In some aspects, the biological samples are obtained every at least, at most, approximately, or exactly 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, or more (or any range or value derivable therein) hours, days, weeks, months, or years during the treatment regime. In some aspects, the level or amount of the biomarkers in each biological sample are compared, which can allow for a determination of the response of the biomarker levels or amounts to the treatment regime. In certain aspects, the response informs whether the treatment regime had clinical significance, including whether the treatment regime treated and / or prevented CA and / or CASH in the patient.

[0018] In certain aspects, a trained model is applied to the amount of biomarker(s) measured in the biological sample obtained from the patient. In some aspects, the trained model is a trained using a leave-one-out cross validation method, a sum of squared error method, an Akaike information criterion method, and / or an error rate method. In some aspects, the model is a multivariate model. In certain aspects, the trained model is trained using a training set of measured biomarkers. In some aspects, the training set of measured biomarkers are biomarkers measured in a population of individuals known to have CA and / or CASH and / or known to not300224353.1 - 5 -have CA and / or CASH. In some aspects, the trained model is applied to predict or determine a response to a therapy administered to the patient.

[0019] The subject may be one that has been diagnosed with CA. In some aspects, the subject has not been diagnosed with CA. The CA may be further defined as CASH, including if associated with symptomatic hemorrhage. The CA may be further defined as non-CASH, including if it is not associated with symptomatic hemorrhage. The associations may be diagnostic if the symptomatic hemorrhage occurred previously. The associations may be prognostic if the hemorrhage develops in the future. The CA may be further defined as aggressive CA. In some aspects, CA is further defined as familial-CA or multifocal-CA or sporadic-CA or solitary-CA. In some aspects, familial CA is further define as familial-CCMl, familial-CCM2, familial-CCM3 or familial-unknown genotype. In some aspects, familial CA is further define as familial-CCM3. The subject may be one that is being treated with a therapy for CA. In some aspects, the subject is not being treated with a therapy for CA. In some aspects, the methods comprise or further comprise administering a therapy for CA. The therapy may comprise or consist of a gene therapy, a B-cell immunomodulation therapy, a statin therapy, a ROCK modulation therapy, surgical excision of a CCM lesion, or combinations thereof. The therapy may also exclude one or more of a gene therapy, a B-cell immunomodulation therapy, a statin therapy, a ROCK modulation therapy, and surgical excision of a CCM lesion.

[0020] The biological sample from the subject may comprise a tissue sample, a blood sample, a whole blood sample, a fractionated sample, a plasma sample, a fecal sample, or a urine sample. In some aspects, the biological sample from the subject comprises a plasma sample from the subject. In some aspects, the biological sample from the subject comprises a sample type described herein.

[0021] Methods for measuring biomarkers are known in the art. For example, the metabolites may be measured by liquid chromatography, mass spectrometry, tandem mass spectrometry, immunoassays (such as an ELISA assay), or by using detection reagents. Also for example, the proteins may be measured by an immunoassay (such as an ELISA assay), mass spectrometry, chromatography, or by using detection reagents. Data analyses can be performed using a complete a lock mass correction using hexakis (1H, 1H, 2H-difluoroethoxy) phosphazene (Synquest Laboratories, Alachua, FL) implemented within the Bruker Data Analysis Software. Feature detection can be done with the open-source software MZmine v2.38 [Pluskal], The unsupervised, differential metabolome between CASH and non-CASH can be performed with PLS-Discriminant Analysis (PLS-DA,) using the R Software. A random permutations analysis on the peak values of the differentially expressed biomarker can be300224353.1 - 6 -performed to estimate the significance of the computed PLS-DA coefficients. Normalized plasma concentrations of each biomarker can be calculated for each patient. Methods for measuring protein levels are known in the art. Protein concentrations can be determined using enzyme-linked immunosorbent assay (ELISA) assays or multiplex electrochemiluminescence immunoassays.

[0022] Also disclosed are systems comprising at least one hardware processor and a non-transitory computer-readable storage medium having stored thereon program code, the program code executable by the at least one hardware processor to: obtain a measured amount of one or more biomarkers in a biological sample obtained from a patient that has received a therapy and apply a trained machine learning model to the measured amount of the one or more biomarkers to predict or determine a response in the patient to the therapy. Also disclosed are systems comprising at least one hardware processor and a non-transitory computer-readable storage medium having stored thereon program code, the program code executable by the at least one hardware processor to: obtain a measured amount of one or more biomarkers in a biological sample obtained from a patient that has received a therapy; train a machine learning model on a training set of measured biomarkers; and apply a trained machine learning model to the measured amount of the one or more biomarkers to predict or determine a response in the patient to the therapy. In some aspects, the training set of measured biomarkers are biomarkers measured in a population of individuals known to have CA and / or CASH and / or known to not have CA and / or CASH. In certain aspects, the biomarkers comprise or consist of tyrosine, acetyl-L-carnitine, arachidonic acid, hexanoylcarnitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, vascular endothelial growth factor- A (VEGF), interleukin (IL)- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, thrombospondin- 1 (TSP1), thrombospondin-2 (TSP2), thrombomodulin (TM), cluster of differentiation- 14 (CD14), c-reactive protein, roundabout guidance receptor 4 (ROBO4), lipopolysaccharide binding protein (LBP), or a combination thereof. In some aspects, one or more of the preceding biomarkers are excluded. In certain aspects, the biomarkers comprise or consist of ROBO4, CD14, TM, and acetyl-L-camitine. In certain aspects, the biomarkers comprise or consist of ROBO4, CD14, and TM. In certain aspects, the biomarkers comprise or consist of endoglin, pipecolic acid, arachidonic acid, and hypoxanthine. In certain aspects, the biomarkers comprise or consist of endoglin, TSP2, and IL-16.

[0023] The kits of the disclosure may comprise, comprise at least, or comprise at most 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29,300224353.1 - 7 -30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, or 100 (or any derivable range therein) detection agents for detecting one or more biomarkers, including any biomarkers disclosed herein such as tyrosine, acetyl-L-carnitine, arachidonic acid, hexanoylcarnitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, vascular endothelial growth factor-A (VEGF), interleukin (IL)- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, thrombospondin- 1 (TSP1), thrombospondin-2 (TSP2), thrombomodulin (TM), cluster of differentiation- 14 (CD14), c-reactive protein, roundabout guidance receptor 4 (ROBO4), lipopolysaccharide binding protein (LBP), or a combination thereof. In some aspects, detection agents for one or more of the preceding biomarkers are specifically excluded from the kit. The detection agents may comprise antibodies, nucleic acid probes, or peptides that are useful for detecting the biomarkers. Kit aspects of the disclosure may comprise one or more negative or positive control samples and / or control detection agents. The kit may comprise or further comprise instructions for use.

[0024] The kit may comprise one or more negative or positive control samples and / or control detection agents. The kit may comprise instructions for use. The kits may comprise reagents for isolation and / or amplification of biomarkers from a biological sample. The kit may exclude reagents for detection of any other biomarkers.

[0025] The kit may comprise reagents for isolation and / or amplification of biomarkers from a biological sample. The biological sample may be a sample from a subject. In some aspects, the kit may excludes reagents for detection of any other biomarkers.

[0026] Also disclosed are methods of measuring biomarkers in a patient having, suspected of having, or diagnosed with having cavernous angioma (CA), methods of measuring biomarkers in a biological sample obtained from a patient, methods of detecting biomarkers in a patient, methods of identifying a change in biomarkers in a patient, methods of detecting biomarkers in a biological sample obtained from a patient.

[0027] In certain aspects, the method comprises one or more steps including measuring the amount of one or more biomarkers from tyrosine, acetyl-L-carnitine, arachidonic acid, hexanoylcarnitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, vascular endothelial growth factor-A (VEGF), interleukin (IL)- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, thrombospondin- 1 (TSP1), thrombospondin-2 (TSP2), thrombomodulin (TM), cluster of300224353.1 - 8 -differentiation-14 (CD14), c-reactive protein, roundabout guidance receptor 4 (ROBO4), lipopolysaccharide binding protein (LBP), or a combination thereof. In certain aspects, the method comprises measuring 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, or 27 biomarkers (or any range derivable therein) from tyrosine, acetyl-L-carnitine, arachidonic acid, hexanoylcamitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, vascular endothelial growth factor- A (VEGF), interleukin (IL)- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, thrombospondin- 1 (TSP1), thrombospondin-2 (TSP2), thrombomodulin (TM), cluster of differentiation- 14 (CD 14), c-reactive protein, roundabout guidance receptor 4 (ROBO4), lipopolysaccharide binding protein (LBP). In certain aspects, one or more of the biomarkers of tyrosine, acetyl-L-carnitine, arachidonic acid, hexanoylcamitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, vascular endothelial growth factor-A (VEGF), interleukin (IL)- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, thrombospondin- 1 (TSP1), thrombospondin-2 (TSP2), thrombomodulin (TM), cluster of differentiation- 14 (CD14), c-reactive protein, roundabout guidance receptor 4 (ROBO4), lipopolysaccharide binding protein (LBP), or a combination thereof are not measured. In certain aspects, one or more of the biomarkers are determined to have a differential level relative to a control. In some aspects, one or more of the biomarkers are determined to have levels not significantly different than the control. In certain aspects, one or more of the biomarkers are determined to have levels not significantly different than the control. In certain aspects, the control comprises the level of the biomarker in samples from subjects identified as non-CASH, low risk, or as not having CA. In certain aspects, the control comprises the level of the biomarker in samples from subjects identified as CASH, high risk, as having CA, or as having aggressive CA. In certain aspects, the CA is further defined as CA with symptomatic hemorrhage (CASH). In certain aspects, the CA is further defined as CA without symptomatic hemorrhage (non-CASH). In certain aspects, the sample from the subject comprises a tissue sample, a blood sample, a whole blood sample, a fractionated sample, a plasma sample, a fecal sample, or a urine sample. In certain aspects, the expression level of no other biomarker in the sample is measured. In certain aspects, the method comprises measuring a protein, including one or more of the proteins of Tables 2, 4, or 5. In some aspects, the method comprises administering a treatment to the patient. The treatment may comprise a gene therapy, a B-cell immunomodulation therapy, a statin therapy, a ROCK modulation therapy, a VEGF inhibitor, surgical excision of a CA lesion, or combinations thereof.300224353.1 - 9 -

[0028] Also disclosed are methods for treating cavernous angioma (CA) or CASH in a patient. The method can comprise one or more steps including administering a therapy to a patient. In some aspects, the patient has been evaluated for one or more biomarkers. In some aspects, the biomarkers including one or more of tyrosine, acetyl-L-camitine, arachidonic acid, hexanoylcarnitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, vascular endothelial growth factor-A (VEGF), interleukin (IL)- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, thrombospondin- 1 (TSP1), thrombospondin-2 (TSP2), thrombomodulin (TM), cluster of differentiation- 14 (CD14), c-reactive protein, roundabout guidance receptor 4 (ROBO4), lipopolysaccharide binding protein (LBP), or a combination thereof are measured in a biological sample from the patient. In certain aspects, the patient has been evaluated for 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, or 27 biomarkers (or any range derivable therein) from tyrosine, acetyl-L-carnitine, arachidonic acid, hexanoylcarnitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, vascular endothelial growth factor-A (VEGF), interleukin (IL)- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, thrombospondin- 1 (TSP1), thrombospondin-2 (TSP2), thrombomodulin (TM), cluster of differentiation- 14 (CD14), c-reactive protein, roundabout guidance receptor 4 (ROBO4), lipopolysaccharide binding protein (LBP), or a combination thereof. In some aspects, one or more of the biomarkers are determined to have a differential level relative to a control. In some aspects, one or more of the biomarkers are determined to have levels not significantly different than the control. In certain aspects, one or more of the biomarkers are determined to have levels not significantly different than the control. In some aspects, the control comprises the level of the biomarker in samples from subjects identified as non-CASH, low risk, or as not having CA. In some aspects, the control comprises the level of the biomarker in samples from subjects identified as CASH prior to sampling or thereafter, or as having CA or not. In some aspects, the measured level of the evaluated biomarkers are input in a statistical model, including a logistic regression or machine learning combination of the weighted levels of one or more biomarkers. In some aspects, a score is calculated from the measured biomarkers.

[0029] In some aspects, the method comprises administering a therapy to a patient that has had one or more biomarkers measured in a biological sample from the patient, and where the measured biomarkers are input into a statistical model, such as a logistical regression, to determine a score. In certain aspects, the therapy is selected or adjusted by the determined300224353.1 - 10 -score. In certain aspects, the score determines whether the patient has responded, will respond, or has a likelihood of responding, to the selected therapy.

[0030] In some aspects, the CA is further defined as CA with symptomatic hemorrhage (CASH). In some aspects, the CA is further defined as CA without symptomatic hemorrhage (non-CASH). In some aspects, the sample from the patient comprises a tissue sample, a blood sample, a whole blood sample, a fractionated sample, a plasma sample, a fecal sample, or a urine sample. In some aspects, the sample from the patient comprises a plasma sample from the patient. In some aspects, the expression level of no other biomarker in the sample was determined. In some aspects, the patient is also evaluated for expression of a protein. In certain aspects, the protein comprises one or more of the proteins of Tables 2, 4, or 5. In some aspects, the treatment comprises a gene therapy, a B-cell immunomodulation therapy, a statin therapy, a ROCK modulation therapy, a VEGF inhibitor, surgical excision of a CA lesion, or combinations thereof.

[0031] Also disclosed are methods for prognosing, diagnosing, or monitoring CA, CASH, or non-CASH in a patient. In certain aspects, the method one or more steps including measuring the amount of one or more biomarkers from tyrosine, acetyl-L-carnitine, arachidonic acid, hexanoylcarnitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, vascular endothelial growth factor-A (VEGF), interleukin (IL)- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, thrombospondin- 1 (TSP1), thrombospondin-2 (TSP2), thrombomodulin (TM), cluster of differentiation- 14 (CD14), c-reactive protein, roundabout guidance receptor 4 (ROBO4), lipopolysaccharide binding protein (LBP), or a combination thereof in a biological sample form the patient. In some aspects, the method comprises measuring 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, or 27 biomarkers (or any range derivable therein) from tyrosine, acetyl-L-carnitine, arachidonic acid, hexanoylcarnitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, vascular endothelial growth factor-A (VEGF), interleukin (IL)- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, thrombospondin- 1 (TSP1), thrombospondin-2 (TSP2), thrombomodulin (TM), cluster of differentiation- 14 (CD14), c-reactive protein, roundabout guidance receptor 4 (ROBO4), lipopolysaccharide binding protein (LBP), or a combination thereof. In certain aspects, the measuring step occurs more than once. In certain aspects, the measuring step is performed after the patient has been administered a therapy. In certain aspects, the measuring step is performed to determine the patient’s response to a therapy. In certain aspects, the monitoring is performed to determine whether a patient has300224353.1 - 11 -or will develop symptomatic hemorrhage. In certain aspects, the monitoring is performed to determine whether a patient has or will develop bleeding. In certain aspects, the monitoring is performed to provide a risk score for determining whether a patient will develop symptomatic hemorrhage. In certain aspects, the patient has received or will receive a therapy, including prior to, during, or after the measuring. In certain aspects, a therapy administering to the patient is adjusted based on the measured biomarker(s).

[0032] In some aspects, the CA is further defined as CA with symptomatic hemorrhage (CASH). In some aspects, the CA is further defined as CA without symptomatic hemorrhage (non-CASH). In some aspects, the sample from the patient comprises a tissue sample, a blood sample, a whole blood sample, a fractionated sample, a plasma sample, a fecal sample, or a urine sample. In some aspects, the sample from the patient comprises a plasma sample from the patient. In some aspects, the expression level of no other biomarker in the sample was determined. In some aspects, the patient is also evaluated for expression of a protein. In certain aspects, the protein comprises one or more of the proteins of Tables 2, 4, or 5. In some aspects, the treatment comprises a gene therapy, a B-cell immunomodulation therapy, a statin therapy, a ROCK modulation therapy, a VEGF inhibitor, surgical excision of a CA lesion, or combinations thereof.

[0033] In some aspects, the expression level of no other biomarker in the sample was measured. In certain aspects, the method comprises measuring a protein. In some aspects, the protein comprises one or more of the proteins of Tables 2, 4, or 5. In some aspects, the method comprises administering a treatment to the patient. In some aspects, the treatment comprises a gene therapy, a B-cell immunomodulation therapy, a statin therapy, a ROCK modulation therapy, a VEGF inhibitor, surgical excision of a CA lesion, or combinations thereof.

[0034] The subject may be a mammal. In some aspects, the subject comprises a laboratory test animal, such as a mouse, rat, rabbit, dog, cat, horse, or pig. In some aspects, the subject is a human. In some aspects, the subject is a human patient.

[0035] Also disclosed are any of the following enumerated Aspects.Aspect 1 includes a method of measuring biomarkers in a patient having, suspected of having, or diagnosed with having cavernous angioma (CA), the method comprising measuring an amount of one or more biomarkers in a biological sample from the patient, wherein the one or more biomarkers comprise tyrosine, acetyl-L-camitine, arachidonic acid, hexanoylcamitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, vascular endothelial growth factor-A (VEGF), interleukin (IL)- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, thrombospondin- 1300224353.1 - 12 -(TSP1), thrombospondin-2 (TSP2), thrombomodulin (TM), cluster of differentiation- 14 (CD 14), c-reactive protein, roundabout guidance receptor 4 (ROBO4), lipopolysaccharide binding protein (LBP), or a combination thereof.Aspect 2 depends upon Aspect 1, wherein the one or more biomarkers comprise ROBO4, CD 14, TM, and acetyl-L-carnitine in the biological sample.Aspect 3 depends upon Aspect 1, wherein the one or more biomarkers comprise ROBO4, CD 14, and TM in the biological sample.Aspect 4 depends upon Aspect 1, wherein the one or more biomarkers comprise endoglin, pipecolic acid, arachidonic acid, and hypoxanthine in the biological sample.Aspect 5 depends upon Aspect 1, wherein the one or more biomarkers comprise endoglin, TSP2, and IL- 16 in the biological sample.Aspect 6 depends upon any one of Aspects 1 to 5, wherein one or more of the biomarkers are determined to have differential levels relative to a control.Aspect 7 depends upon Aspect 6, wherein one or more of the biomarkers are determined to have levels significantly different than the control.Aspect 8 depends upon Aspect 6, wherein one or more of the biomarkers are determined to have levels not significantly different than the control.Aspect 9 depends upon any one of Aspects 6 to 8, wherein the control comprises the level of the biomarker(s) in samples from subjects identified as non-CASH, low risk, as not having CA, or as having non-aggressive CA.Aspect 10 depends upon any one of Aspects 6 to 9, wherein the control comprises the level of the biomarker(s) in samples from subjects identified as CASH, high risk, as having CA, or as having aggressive CA.Aspect 11 depends upon any one of Aspects 6 to 8, wherein the control comprises the level of the biomarker(s) in a second biological sample obtained from the patient prior to the measuring.Aspect 12 depends upon any one of Aspects 6 to 11, wherein the measuring is performed on two or more biological samples obtained from the patient.Aspect 13 depends upon Aspect 12, wherein the two or more biological samples are obtained at different times.Aspect 14 depends upon any one of Aspects 1 to 13, wherein the CA is further defined as CA with symptomatic hemorrhage (CASH).Aspect 15 depends upon any one of Aspects 1 to 14, wherein the CA is further defined as CA without symptomatic hemorrhage (non-CASH).300224353.1 - 13 -Aspect 16 depends upon any one of Aspects 1 to 15, wherein the sample from the subject comprises a tissue sample, a blood sample, a whole blood sample, a fractionated sample, a plasma sample, a fecal sample, or a urine sample.Aspect 17 depends upon any one of Aspects 1 to 16, wherein the sample from the subject comprises a plasma sample from the subject.Aspect 18 depends upon any one of Aspects 1 to 17, wherein the expression level of no other biomarker in the sample is measured.Aspect 19 depends upon any one of Aspects 1 to 18, further comprising administering a treatment to the patient.Aspect 20 depends upon Aspect 19, wherein the treatment comprises a gene therapy, a B-cell immunomodulation therapy, surgical excision of a CCM lesion, or combinations thereof. Aspect 21 includes a method for treating cavernous angioma (CA) or CASH in a patient, the method comprising administering a therapy to a subject that has been evaluated for one or more biomarkers in a biological sample from the patient, wherein the one or more biomarkers comprise tyrosine, acetyl-L-camitine, arachidonic acid, hexanoylcarnitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, VEGF, IL- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, TSP1, TSP2, TM, CD14, c-reactive protein, ROBO4, LBP, or a combination thereof.Aspect 22 depends upon Aspect 21, wherein the one or more biomarkers comprise ROBO4, CD 14, TM, and acetyl-L-carnitine in the biological sample.Aspect 23 depends upon Aspect 21, wherein the one or more biomarkers comprise ROBO4, CD 14, and TM in the biological sample.Aspect 24 depends upon Aspect 21, wherein the one or more biomarkers comprise endoglin, pipecolic acid, arachidonic acid, and hypoxanthine in the biological sample.Aspect 25 depends upon Aspect 21, wherein the one or more biomarkers comprise endoglin, TSP2, and IL- 16 in the biological sample.Aspect 26 depends upon any one of Aspects 21 to 25, wherein one or more of the biomarkers are determined to have a differential level relative to a control.Aspect 27 depends upon Aspect 26, wherein one or more of the biomarkers are determined to have levels significantly different than the control.Aspect 28 depends upon Aspect 26, wherein one or more of the biomarkers are determined to have levels not significantly different than the control.300224353.1 - 14 -Aspect 29 depends upon any one of Aspects 26 to 28, wherein the control comprises the level of the biomarker in samples from subjects identified as non-CASH, low risk, as not having CA, or as having non-aggressive CA.Aspect 30 depends upon any one of Aspects 26 to 29, wherein the control comprises the level of the biomarker in samples from subjects identified as CASH, high risk, as having CA, or as having aggressive CA.Aspect 31 depends upon any one of Aspects 26 to 28, wherein the control comprises the level of the biomarker(s) in a second biological sample obtained from the patient prior to the measuring.Aspect 32 depends upon any one of Aspects 26 to 31, wherein the measuring is performed on two or more biological samples obtained from the patient.Aspect 33 depends upon Aspect 32, wherein the two or more biological samples are obtained at different times.Aspect 34 depends upon any one of Aspects 21 to 33, wherein the CA is further defined as CA with symptomatic hemorrhage (CASH).Aspect 35 depends upon any one of Aspects 21 to 34, wherein the CA is further defined as CA without symptomatic hemorrhage (non-CASH).Aspect 36 depends upon any one of Aspects 21 to 35, wherein the sample from the subject comprises a tissue sample, a blood sample, a whole blood sample, a fractionated sample, a plasma sample, a fecal sample, or a urine sample.Aspect 37 depends upon any one of Aspects 21 to 36, wherein the sample from the subject comprises a plasma sample from the subject.Aspect 38 depends upon any one of Aspects 21 to 37, wherein the expression level of no other biomarker in the sample was evaluated.Aspect 39 depends upon any one of Aspects 21 to 38, wherein the treatment comprises a gene therapy, a B-cell immunomodulation therapy, surgical excision of a CCM lesion, or combinations thereof.Aspect 40 includes a method for prognosing, diagnosing, or monitoring CA, aggressive CA, non-aggressive CA, CASH or non-CASH, familial CA, or sporadic CA in a subject, the method comprising measuring the amount of one or more biomarkers in a biological sample from the patient, wherein the one or more biomarkers comprise tyrosine, acetyl-L-carnitine, arachidonic acid, hexanoylcamitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, VEGF,300224353.1 - 15 -IL-10, IL-ip, IL-16, endoglin, angiopoietin-1, angiopoietin-2, TSP1, TSP2, TM, CD14, c-reactive protein, ROBO4, LBP, or a combination thereof.Aspect 41 depends upon Aspect 40, wherein the one or more biomarkers comprise ROBO4, CD 14, TM, and acetyl-L-carnitine in the biological sample.Aspect 42 depends upon Aspect 40, wherein the one or more biomarkers comprise ROBO4, CD 14, and TM in the biological sample.Aspect 43 depends upon Aspect 40, wherein the one or more biomarkers comprise endoglin, pipecolic acid, arachidonic acid, and hypoxanthine in the biological sample.Aspect 44 depends upon Aspect 40, wherein the one or more biomarkers comprise endoglin, TSP2, and IL- 16 in the biological sample.Aspect 45 depends upon any one of Aspects 40 to 44, wherein one or more of the biomarkers are determined to have a differential level relative to a control.Aspect 46 depends upon Aspect 45, wherein one or more of the biomarkers are determined to have levels significantly different than the control.Aspect 47 depends upon Aspect 45, wherein one or more of the biomarkers are determined to have levels not significantly different than the control.Aspect 48 depends upon any one of Aspects 45 to 47, wherein the control comprises the level of the biomarker in samples from subjects identified as non-CASH, low risk, as not having CA, or as having non-aggressive CA.Aspect 49 depends upon any one of Aspects 45 to 48, wherein the control comprises the level of the biomarker in samples from subjects identified as CASH, high risk, as having CA, or as having aggressive CA.Aspect 50 depends upon any one of Aspects 45 to 47, wherein the control comprises the level of the biomarker(s) in a second biological sample obtained from the patient.Aspect 51 depends upon Aspect 50, wherein the second biological sample is obtained at a different time from obtaining the biological sample.Aspect 52 depends upon any one of Aspects 40 to 51, wherein the CA is further defined as CA with symptomatic hemorrhage (CASH).Aspect 53 depends upon any one of Aspects 40 to 52, wherein the CA is further defined as CA with symptomatic hemorrhage (CASH).Aspect 54 depends upon any one of Aspects 40 to 52, wherein the CA is further defined as CA without symptomatic hemorrhage (non-CASH).300224353.1 - 16 -Aspect 55 depends upon any one of Aspects 40 to 54, wherein the sample from the subject comprises a tissue sample, a blood sample, a whole blood sample, a fractionated sample, a plasma sample, a fecal sample, or a urine sample.Aspect 56 depends upon any one of Aspects 40 to 55, wherein the sample from the subject comprises a plasma sample from the subject.Aspect 57 depends upon any one of Aspects 40 to 56, wherein the expression level of no other biomarker in the sample was measured.Aspect 58 depends upon any one of Aspects 40 to 57, further comprising administering a treatment to the patient.Aspect 59 depends upon Aspect 58, wherein the treatment comprises a gene therapy, a B-cell immunomodulation therapy, surgical excision of a CCM lesion, or combinations thereof. Aspect 60 includes a kit comprising detection agents for determining levels of one or more biomarkers, wherein the biomarkers comprise tyrosine, acetyl-L-carnitine, arachidonic acid, hexanoylcarnitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, VEGF, IL- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, TSP1, TSP2, TM, CD14, c-reactive protein, ROBO4, LBP, or a combination thereof.Aspect 61 depends upon Aspect 60, wherein the one or more biomarkers comprise ROBO4, CD 14, TM, and acetyl-L-carnitine.Aspect 62 depends upon Aspect 60, wherein the one or more biomarkers comprise ROBO4, CD 14, and TM.Aspect 63 depends upon Aspect 60, wherein the one or more biomarkers comprise endoglin, pipecolic acid, arachidonic acid, and hypoxanthine.Aspect 64 depends upon Aspect 60, wherein the one or more biomarkers comprise endoglin, TSP2, and IL- 16.Aspect 65 depends upon any one of Aspects 60 to 64, wherein the kit further comprises one or more negative or positive control samples and / or control detection agents.Aspect 66 depends upon any one of Aspects 60 to 65, wherein the kit further comprises instructions for use.Aspect 67 depends upon any one of Aspects 60 to 66, wherein the kit comprises reagents for isolation and / or amplification of biomarkers from a biological sample.Aspect 68 depends upon Aspect 67, wherein the biological sample comprises a tissue sample, a blood sample, a whole blood sample, a fractionated sample, a plasma sample, a fecal sample, or a urine sample.300224353.1 - 17 -Aspect 69 depends upon Aspect 67, wherein the biological sample comprises a plasma sample.Aspect 70 depends upon any one of Aspects 67 to 69, wherein the biological sample is from a subject.Aspect 71 depends upon any one of Aspects 60 to 70, wherein the subject has been diagnosed with CA.Aspect 72 depends upon any one of Aspects 60 to 71, wherein the kit excludes reagents for detection of any other biomarkers.

[0036] Treatment” or treating may refer to any treatment of a disease in a mammal, including: (i) preventing the disease, that is, causing the clinical symptoms of the disease not to develop by administration of a protective composition prior to the induction of the disease; (ii) suppressing the disease, that is, causing the clinical symptoms of the disease not to develop by administration of a protective composition after the inductive event but prior to the clinical appearance or reappearance of the disease; (iii) inhibiting the disease, that is, arresting the development of clinical symptoms by administration of a protective composition after their initial appearance; and / or (iv) relieving the disease, that is, causing the regression of clinical symptoms by administration of a protective composition after their initial appearance. In some aspects, the treatment may exclude prevention of the disease.

[0037] Throughout this application, the term “about” is used according to its plain and ordinary meaning in the area of cell and molecular biology to indicate that a value includes the standard deviation of error for the device or method being employed to determine the value.

[0038] The use of the word “a” or “an” when used in conjunction with the term “comprising” may mean “one,” but it is also consistent with the meaning of “one or more,” “at least one,” and “one or more than one.”

[0039] As used herein, the terms “or” and “and / or” are utilized to describe multiple components in combination or exclusive of one another. For example, “x, y, and / or z” can refer to “x” alone, “y” alone, “z” alone, “x, y, and z,” “(x and y) or z,” “x or (y and z),” or “x or y or z.” It is specifically contemplated that x, y, or z may be specifically excluded from an embodiment or aspect.

[0040] The words “comprising” (and any form of comprising, such as “comprise” and “comprises”), “having” (and any form of having, such as “have” and “has”), “including” (and any form of including, such as “includes” and “include”), “characterized by” (and any form of including, such as “characterized as”), or “containing” (and any form of containing, such as “contains” and “contain”) are inclusive or open-ended and do not exclude additional, unrecited elements or method steps.300224353.1 - 18 -

[0041] The compositions and methods for their use can “comprise,” “consist essentially of,” or “consist of’ any of the ingredients or steps disclosed throughout the specification. The phrase “consisting of’ excludes any element, step, or ingredient not specified. The phrase “consisting essentially of’ limits the scope of described subject matter to the specified materials or steps and those that do not materially affect its basic and novel characteristics. It is contemplated that embodiments and aspects described in the context of the term “comprising” may also be implemented in the context of the term “consisting of’ or “consisting essentially of.”

[0042] As used herein, the term “subject” may be used interchangeably with “patient”. In certain aspects, the patient is a human patient. In certain aspects, the patient is a patient enrolled in a clinical trial.

[0043] Any method in the context of a therapeutic, diagnostic, or physiologic purpose or effect may also be described in “use” claim language such as “Use of’ any compound, composition, or agent discussed herein for achieving or implementing a described therapeutic, diagnostic, or physiologic purpose or effect.

[0044] Use of the one or more sequences or compositions may be employed based on any of the methods described herein. Other aspects are discussed throughout this application. Any embodiment or aspect discussed with respect to one aspect of the disclosure applies to other aspects of the disclosure as well and vice versa.

[0045] It is specifically contemplated that any limitation discussed with respect to one embodiment or aspect of the invention may apply to any other embodiment or aspect of the invention. Furthermore, any composition of the invention may be used in any method of the invention, and any method of the invention may be used to produce or to utilize any composition of the invention. Aspects of an embodiment set forth in the Examples are also aspects that may be implemented in the context of aspects discussed elsewhere in a different Example or elsewhere in the application, such as in the Summary of Invention, Brief Description of the Drawings, Detailed Description, and Claims.

[0046] Other obj ects, features and advantages of the present invention will become apparent from the following detailed description. It should be understood, however, that the detailed description and the specific examples, while indicating specific embodiments and aspects of the invention, are given by way of illustration only, since various changes and modifications within the spirit and scope of the invention will become apparent to those skilled in the art from this detailed description.300224353.1 - 19 -BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The following drawings form part of the present specification and are included to further demonstrate certain aspects of the present invention. The invention may be better understood by reference to one or more of these drawings in combination with the detailed description of specific embodiments presented herein.

[0048] FIG. 1. Weighted combination of 3 circulating metabolites distinguish diagnostic CASH and non-CASH patients with 70 / 95% of specificity / sensitivity (Accuracy = 83%)

[0049] FIG. 2. A combination of protein and metabolite levels improved diagnostic association of CASH and non-CASH patients to 80 / 80% of specificity / sensitivity (Accuracy = 80%).

[0050] FIGS. 3A-3B. Combinations of plasma level change of proteins or metabolites vs. relative change in mean lesional quantitative susceptibility mapping (QSM). FIG. 3A shows model derived from relative change in roundabout guidance receptor 4 (ROBO4), cluster of differentiation- 14 (CD 14) and thrombomodulin (TM) plasma level versus relative change in mean QSM; area under the curve (AUC)=95.6° / > [90.5%; 100%]; Sn=95.0% and Sp=81.0% with a threshold value of 0.33 (p= 8.2xl0-11). FIG. 3B shows model derived from relative change in acetyl. L. carnitine and lysophosphatidylethanolamine (LPE) 18:0 plasma level predicting the categorical change in mean QSM; AUC=95.9% [ [91.5%; 100%] (^=1.5xl0’n), Sn=95.0% and Sp=87.0% with a threshold value of 0.34 (j?=1.5xl0‘11). Symptomatic hemorrhage (SH) cases are noted in in red and asymptomatic change (AC) cases in blue.

[0051] FIGS. 4A-4B. Combinations of plasma level change of proteins and metabolites vs. relative and in mean quantitative susceptibility mapping (QSM). FIG. 4A shows model derived from relative change in seven proteins [lipopolysaccharide binding protein (LBP), vascular endothelial growth factor-A (VEGF), interleukin (IL)ip, thrombomodulin (TM), toll-like receptor 4 (TLR4), IL16 and thrombospondin-2 (TSP2)] and one metabolite (methionine) plasma level predicting the relative change in mean QSM with Sn=86.0%; Sp=89.0%; p=0.66, R2=0.44 and sum of squared error under leave-one-out cross validation of 2.89 (p=0.002. FIG.4B shows model derived from relative change in three proteins (roundabout guidance receptor 4 [ROBO4], TM and cluster of differentiation- 14 [CD14]) and one metabolite (acetyl. L. carnitine) plasma level predicting the categorical change in mean QSM; area under the curve (AUC)=98.9% [97.2%; 100%]; Sn=100%; Sp= 89.0% with a threshold value of 0.33 (j?=3.1xl0‘13). Symptomatic hemorrhage (SH) cases are noted in in red and asymptomatic change (AC) cases in blue.300224353.1 - 20 -

[0052] FIGS. 5A-5B. Combinations of plasma level change of proteins or metabolites vs. relative change in mean dynamic contrast enhanced quantitative perfusion (DCEQP). FIG. 5A shows model derived from absolute change in endoglin, thrombospondin-2 (TSP2) and interleukin (IL) 16 plasma level predicting the categorical change in mean DCEQP; AUC=98.2° / o [94.7%; 100%]; Sn=97.0% and Sp=97.0% for a threshold value of 0.11 (p=8.8xl0‘15). FIG. 5B shows model derived from relative change in piperine, arachidonic acid and hypoxanthine plasma level predicting the categorical change in mean DCEQP; area under the curve (AUC)=99.1% [97.5%; 100%]; Sn=100% and Sp=93.0% for a threshold value of 0.43 ( / ?=1.4xl0‘15). One of the two symptomatic hemorrhage (SH) cases (in red) and one of the three asymptomatic change (AC) cases (in blue) were detected by the categorical change in mean DCEQP (ADCEQP>0.40).

[0053] FIG. 6. Combinations of plasma level change of proteins and metabolites vs. categorical change in mean dynamic contrast enhanced quantitative perfusion (DCEQP). Model derived from relative change in one protein (endoglin) and three metabolites (pipecolic acid, arachidonic acid, hypoxanthine) predicting the categorical change in mean DCEQP with area under the curve (AUC)=99.9° / o [99.6%; 100%]; Sn=97.0%> and Sp=100% considering a threshold value of 0.23 (p=4.1x10⁻17). One of the two symptomatic hemorrhage (SH) cases (in red) and one of the three asymptomatic change (AC) cases (in blue) were detected by the categorical change in mean DCEQP (ADCEQP>0.40).

[0054] FIGS. 7A-7B. Biological significance of metabolites and proteins correlating with quantitative susceptibility mapping (QSM) and dynamic contrast enhanced quantitative perfusion (DCEQP). Proteins involved in QSM and DCEQP enriched pathways queried by DAVID Knowledgebase (p< 0.05, Bonferroni corrected). Metabolites involved in QSM and DCEQP enriched pathways queried by CTD (p< 0.01, Bonferroni corrected). Pathways classified into categories of cell proliferation, apoptosis and oxidative stress, vascular processes, inflammation / immune response, permeability / adhesion, and metabolism. FIG. 7A shows proteome pathways connected to metabolites correlated with QSM (methionine, acetyl-L-camitine, lysophosphatidylethanolamine [LPE] 18.0) and proteins (roundabout guidance receptor 4 [ROBO4], toll-like receptor 4 [TLR4], thrombomodulin [TM], vascular endothelial growth factor-A [VEGF], lipopolysaccharide binding protein [LBP], interleukin [IL] 16, ILip, thrombospondin- 1 [TSP1], cluster of differentiation- 14 [CD14]). FIG. 7B shows proteome pathways connected to the metabolites correlated with DCEQP (hypoxanthine, arachidonic acid, pipecolic acid, piperine) and proteins (thrombospondin-2 [TSP2], IL-16, endoglin).300224353.1 - 21 -

[0055] FIGS. 8A-8H. QSM and DCEQP absolute and percent changes detecting symptomatic hemorrhage (SH) in CASH Trial Readiness (TR) Project1. FIG. 8A shows absolute change of mean lesional QSM, which was positive in all 7 SH cases, 7 of 10 asymptomatic cases (AC) and in 68 out of 127 lesions with no clinical events detected. FIG.8B shows absolute change of mean lesional QSM derived from the TR Project documented an area under the curve (AUC) of 74.0% for detecting symptomatic hemorrhage with a sensitivity of 100% and a specificity of 60.0% (p <1 O’4). FIG. 8C shows percent change of mean lesional QSM of 6% f or more was reported in all 7 SH cases, in 7 of 10 AC cases and in 51 out of 127 lesions with no clinical events detected. FIG. 8D shows percent change of mean lesional QSM in the TR Project, which had an AUC of 75.0% for detecting SH with a sensitivity of 100% and a specificity of 57.0% (p < 1 O’4). FIG. 8E shows absolute change of mean lesional DCEQP, which was positive in 6 of 8 SH cases, 5 of 9 AC and in 61 out of 117 lesions with no clinical events detected. FIG. 8F shows absolute change of mean lesional DCEQP which reported an AUC of 73.0% for detecting SH with a sensitivity of 63.0% and a specificity of 79.0% (p=0.02). FIG. 8G shows percent change of mean lesional DCEQP of 40%* or more, which was reported in 6 of 8 SH cases, 3 of 9 AC and in 43 out of 117 lesions with no clinical events detected. FIG.8H shows percent change of mean lesional DCEQP which had an AUC of 73.0% for detecting SH with a sensitivity of 63.0% and a specificity of 79.0% (p=0.02). * Categorical threshold biomarker events defined as> 6% yearly change in mean lesional QSM and >40% annual change in mean lesional DCEQP.

[0056] FIGS. 9A-9B. Combinations of proteins relative and absolute changes vs. relative and absolute change in mean QSM. FIG. 9A shows combination of relative change in proteins levels reflecting the relative change in mean QSM with highest documented correlation p =0.64; R.2=Q.41(^=0.003); / 7=60.0%; 5 / 2=94.0% and sum of squared error (SSE) under LOOCV of 3.06: QSM%~0.32*[LBP%] +0.30*[endoglin%] +0.1*[VEGF%] +0.04*[TM%] +9.8xl0-3*[TLR4%]+5.6xl0~4*[IL16%] +2.3x10~5*[TSP1%] -O.15*[IL1B%] -0.15. FIG. 9B shows combination of absolute change in proteins levels reflecting the absolute change in mean QSM with highest documented correlation of p=0.26; 52=0.07 (p=0.047); 5w=33.0%; 5 / 2=100% and SSE under LOOCV of 3.88: AQSM ~ -2.9xlO~8*[ATSP 1] -8.7xl0~5*[AROBO4]-2.0xl0~4*[ATLR4]-5.4*[AIL1B] -0.07.

[0057] FIG. 10. Change in proteins plasma levels vs. categorical change in mean QSM.Combination of absolute change in proteins plasma levels reflects the categorical change in300224353.1 - 22 -mean lesional QSM>6% with 86.0% / 85.0% sensitivity / specificity [area under the curve (AUC) =90.2% (78.8%; 100%); P= 1.2xl0’8].

[0058] FIG. 11. Absolute change in acetyl. L. carnitine vs. absolute change in mean QSM.Absolute change in acetyl. L. carnitine correlation with absolute change in mean QSM with p=0.43; R2=Q.18 (p=0.001); 5w=32.0%; 5 / 2=81.0% and a sum of squared error under LOOCV of 4.23: Z1QS / V7~ 0.07 * [AacetyLL.carnitine]-0.03.

[0059] FIG. 12. Relative change in metabolites plasma levels vs. relative change in mean QSM. Composite model of relative change in metabolite levels reflecting the relative change in mean QSM with highest documented correlation p=0.20; 52=0.04 (p=0.532); 5w=15.0%; 5 / 2=79.0% and sum of squared error under LOOCV of 8.92: QSM% ~ -0.10* [acetyl. L.carnitine%] -0.01*[phenylacetylglutamine%]- 0.02* [piperine%]+ 0.01.

[0060] FIG. 13. Absolute change in metabolites vs. categorical change in mean QSM.Combination of absolute change in metabolites levels reflecting the categorical change in mean QSM with highest AUC= 80.5% [69.6%; 91.5%]; Sn=75.0% and Sp=72.0% for a cutoff value of 0.35 (p=0.004): biQSM ~ 0.08* [AacetyLL.carnitine] -0.32*[Apipecolic acid] -0.68.

[0061] FIGS. 14A-14B. Combinations of proteins relative and absolute changes vs. relative and absolute change in mean DCEQP. FIG. 14A shows combination of relative change in proteins levels reflecting the relative change in mean DCEQP with highest documented correlation p =0.25; 52=0.04 (p=0.39); Sn= 100%; Sp=24.0% and sum of squared error (SSE) under LOOCV of 1.69: DCEQP%~-O.3O*[IL1B%] + 0.07*[TLR4%] +0.82. FIG. 14B shows combination of absolute change in proteins levels reflecting the absolute change in mean DCEQP with highest documented correlation p =0.06; R2=0.11 (p=0.22); 5w=96.0%; Sp=39.0% and SSE under LOOCV of 5.31: DCEQP~ -6.34*[AIL1B] -3.92xlO~8*[ATSP 1] + 0.01*[ATSP2]-1.0xl0~4*[AROBO4] +0.07.

[0062] FIG. 15. Relative change in proteins vs. categorical change in mean DCEQP. Combination of relative change in proteins levels reflecting the categorical change in mean DCEQP with highest AUC=92.2% [85.9%; 98.5%]; Sn=82.0% and Sp=81.0% (p=2.7xlO-10) for a cutoff value of 0.53: biDCEQP~ 2.0*[endoglin%] -1.73*[VEGF%] +0.6*[IL16%] -0.86.

[0063] FIGS. 16A-16B. Combinations of metabolites relative and absolute changes vs. relative and absolute change in mean DCEQP. FIG. 16A shows combination of relative change in metabolites levels reflecting the relative change in mean DCEQP with the highest documented correlation p =0.15; 52=0.01 (p=0.86); Sn=100%; Sp=8.0% and sum of squared error (SSE) under LOOCV of 45.48: DCEQP% ~ -0.06*[hypoxanthine%] -300224353.1 - 23 -0.09* [piperine%]+ 0.90. FIG. 16B shows combination of absolute change in metabolites levels reflecting the absolute change in mean DCEQP with the highest documented correlation p= -0.20; R2=0.04 (p=0.31); Sn=93.0%; Sp=21.0% and SSE under LOOCV of 6.24: ADCEQP~ 0.12*[ALPE 18:0] +0.01*[Apiperine ]+0.05.

[0064] FIG. 17. Absolute change in metabolites levels vs. categorical change in mean DCEQP. Combination of absolute change in metabolites levels reflecting the categorical change in mean DCEQP with highest AUC=98.0% [95.4%; 100%] reported high Sn=96.0%; Sp=91.0% for a cutoff value of 0.34 ( / ?=1.4xl0‘15): biDCEQP ~ 0.30*[ALPE 18.0] -0.17*[Apiperine] - 0.04*[Apipecolic acid] -0.73.

[0065] FIGS. 18A-18B. Combinations of proteins and metabolites relative and absolute changes vs. relative and absolute change in mean DCEQP. FIG. 18A shows combination of relative change in proteins and metabolites levels reflecting the relative change in mean DCEQP with highest d correlation p =0.39; R2= -0.01 ( =0.46); 5w=100%; Sp=25.0% and sum of squared error (SSE) under LOOCV of 53.07: DCEQP% ~ -0.27*[ROBO4%] -O.34*[IL1B%] -0.48*[acetyl.L.carnitine%]+ 1.47. FIG. 18B shows combination of absolute change in proteins and metabolites levels reflecting the absolute change in mean DCEQP with highest documented correlation p =0.08; R2=0.03 ( / ?=0.28); Sn=96.0% 5 / 1=46.0% and SSE under LOOCV of 5.25: ADCEQP~ 0.01*[ATSP2] -1.1x10-4*[ΔROBO4] -1.43*[AIL1B]-0.10*[Ahypoxanthine] -4.07xl0~8*[ATSP 1] +0.08.DETAILED DESCRIPTION OF THE INVENTIONI. Measuring Biomarkers

[0066] Biomarkers may be measured using any technique or method described herein. In some aspects, proteins are measured using immunoassay -based methods. These include single-plex and multiplex sandwich assays such as ELISA, electrochemiluminescence immunoassay, bead-based immunoassays (e.g., Luminex), proximity extension assays, and aptamer- or DNA-barcoded antibody platforms. Assays may be configured for absolute quantification using a standard curve prepared from purified calibrators spanning the dynamic range of interest, or for relative quantification using reference controls. Samples and standards may be run in replicate. Signal may be detected by colorimetry, fluorescence, chemiluminescence, or electrochemiluminescence, and concentrations are interpolated from the calibration model (e.g., 4- or 5-parameter logistic fit).300224353.1 - 24 -

[0067] In other aspects, proteins are measured by mass spectrometry. Proteins may be denatured, reduced, alkylated, and digested (e.g., with trypsin) to generate peptides. Targeted quantification may be performed using multiple / selected reaction monitoring (MRM / SRM) or parallel reaction monitoring (PRM) on triple quadrupole or high-resolution instruments. Stable isotope-labeled internal standards (e.g., AQUA peptides) may be spiked into each sample prior to digestion or prior to LC-MS injection to enable absolute quantitation. Untargeted proteomics may be performed using data-dependent acquisition (DDA) or data-independent acquisition (DIA, e.g., SWATH) to profile a broad set of proteins. Peptides may be separated by liquid chromatography and analyzed by MS / MS. In some aspects, proteins are identified by searching spectral data against appropriate databases with control of false discovery rate.

[0068] Alternative or complementary methods may include Western blotting, capillary immunoassay, immunohistochemistry or immunofluorescence on fixed tissue, surface plasmon resonance, microarrays, and biosensors. Post-translational modifications (e.g., phosphorylation, glycosylation) may be assessed using modification-specific antibodies, enrichment strategies, or diagnostic fragment ions.

[0069] In certain aspects, metabolites are measured by chromatography-coupled mass spectrometry or immune-assays. Liquid chromatography-mass spectrometry (LC-MS or LC-MS / MS) utilizing reversed-phase or hydrophilic interaction chromatography (HILIC) may be employed for polar and nonpolar metabolites. Gas chromatography-mass spectrometry (GC-MS) may be used for volatile or derivatized metabolites (e.g., silylated sugars, organic acids). Nuclear magnetic resonance (NMR) spectroscopy can be used for absolute or relative quantification of abundant metabolites without derivatization. Assays may be operated in targeted mode using predefined transitions or exact masses for specific analytes, or in untargeted mode to profile a broad range of small molecules.

[0070] For targeted metabolomics, stable isotope-labeled internal standards corresponding to each analyte or to representative chemical classes may be added to each sample prior to extraction to enable correction for matrix effects and recovery. Calibration curves may be prepared by spiking known concentrations of analytes into stripped matrix or surrogate matrix across the expected physiological range. For untargeted metabolomics, feature detection, alignment, and annotation may be performed using spectral libraries and retention indices, with putative identifications confirmed using authentic standards when available.

[0071] In some aspects, biomarker concentrations are reported as absolute quantities (e.g., ng / mL or pM) based on calibration against standards. In other aspects, biomarker levels are expressed relative to a reference (e.g., fold-change to a control sample, z-scores standardized300224353.1 - 25 -to a healthy population, fold-change relative to a different biological sample). Biomarker measurements from multiple analytes can be combined into a composite index or algorithmic score by applying predetermined weights, logistic models, or machine-learning classifiers. Scores may be scaled to a defined range and threshold(s) established for clinical interpretation II. ROC analysis

[0072] In statistics, a receiver operating characteristic (ROC), or ROC curve, is a graphical plot that illustrates the performance of a binary classifier system as its discrimination threshold is varied. ROC analysis may be applied to determine a cut-off value or threshold setting of biomarker expression, such as the canonical value described herein. For example, patients with biological samples determined to have biomarker expression value above a certain cut-off threshold but below a higher cut-off threshold may be determined to have endometriosis. Patients with biological samples determined to have a biomarker expression level that surpasses the cut-off threshold may be determined to have a disease or condition such as multiple sclerosis. The curve is created by plotting the true positive rate against the false positive rate at various threshold settings. (The true-positive rate is also known as sensitivity in biomedical informatics, or recall in machine learning. The false-positive rate is also known as the fall-out and can be calculated as 1 - specificity). The ROC curve is thus the sensitivity as a function of fall-out. In general, if the probability distributions for both detection and false alarm are known, the ROC curve can be generated by plotting the cumulative distribution function (area under the probability distribution from -infinity to + infinity) of the detection probability in the y-axis versus the cumulative distribution function of the false-alarm probability in x-axis.

[0073] ROC analysis provides tools to select possibly optimal models and to discard suboptimal ones independently from (and prior to specifying) the cost context or the class distribution. ROC analysis is related in a direct and natural way to cost / benefit analysis of diagnostic decision making.

[0074] The ROC curve was first developed by electrical engineers and radar engineers during World War II for detecting enemy objects in battlefields and was soon introduced to psychology to account for perceptual detection of stimuli. ROC analysis since then has been used in medicine, radiology, biometrics, and other areas for many decades and is increasingly used in machine learning and data mining research.

[0075] The ROC is also known as a relative operating characteristic curve, because it is a comparison of two operating characteristics (TPR and FPR) as the criterion changes. ROC analysis curves are known in the art and described in Metz CE (1978) Basic principles of ROC300224353.1 - 26 -analysis. Seminars in Nuclear Medicine 8:283-298; Youden WJ (1950) An index for rating diagnostic tests. Cancer 3:32-35; Zweig MH, Campbell G (1993) Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clinical Chemistry 39:561-577; and Greiner M, Pfeiffer D, Smith RD (2000) Principles and practical application of the receiver-operating characteristic analysis for diagnostic tests. Preventive Veterinary Medicine 45:23-41, which are herein incorporated by reference in their entirety. A ROC analysis may be used to create cut-off values for prognosis and / or diagnosis purposes.III. Sample Preparation

[0076] In certain aspects, methods involve obtaining a sample from a subject. The methods of obtaining provided herein may include methods of biopsy such as fine needle aspiration, core needle biopsy, vacuum assisted biopsy, incisional biopsy, excisional biopsy, punch biopsy, shave biopsy or skin biopsy. The sample may be obtained from any source including but not limited to blood, serum, plasma, sweat, hair follicle, buccal tissue, tears, menses, feces, or saliva. In certain aspects of the current methods, any medical professional such as a doctor, nurse or medical technician may obtain a biological sample for testing. Yet further, the biological sample can be obtained without the assistance of a medical professional.

[0077] A sample may include but is not limited to, tissue, cells, or biological material from cells or derived from cells of a subject. The biological sample may be a heterogeneous or homogeneous population of cells or tissues. The biological sample may be obtained using any method known to the art that can provide a sample suitable for the analytical methods described herein. The sample may be obtained by non-invasive methods including but not limited to: drawing blood by phlebotomy, scraping of the skin or cervix, swabbing of the cheek, saliva collection, urine collection, feces collection, collection of menses, tears, or semen.

[0078] The sample may be obtained by methods known in the art. In certain aspects the samples are obtained by biopsy. In other aspects the sample is obtained by swabbing, endoscopy, scraping, phlebotomy, or any other methods known in the art. In some cases, the sample may be obtained, stored, or transported using components of a kit of the present methods. In some cases, multiple samples, such as multiple plasma or serum samples may be obtained for diagnosis by the methods described herein. In other cases, multiple samples, such as one or more samples from one tissue type (for example ovaries or related tissues) and one or more samples from another specimen (for example serum or plasma) may be obtained for diagnosis by the methods. Samples may be obtained at different times are stored and / or analyzed by different methods. For example, a sample may be obtained and analyzed by routine staining methods or any other cytological analysis methods.300224353.1 - 27 -

[0079] In some aspects the biological sample may be obtained by a physician, nurse, or other medical professional such as a medical technician, endocrinologist, cytologist, phlebotomist, radiologist, or a pulmonologist. The medical professional may indicate the appropriate test or assay to perform on the sample. In certain aspects a molecular profiling business may consult on which assays or tests are most appropriately indicated. In further aspects of the current methods, the patient or subject may obtain a biological sample for testing without the assistance of a medical professional, such as obtaining a whole blood sample, a urine sample, a fecal sample, a buccal sample, or a saliva sample.

[0080] In other cases, the sample is obtained by an invasive procedure including but not limited to: biopsy, needle aspiration, blood draw, endoscopy, or phlebotomy. The method of needle aspiration may further include fine needle aspiration, core needle biopsy, vacuum assisted biopsy, or large core biopsy. In some aspects, multiple samples may be obtained by the methods herein to ensure a sufficient amount of biological material.

[0081] General methods for obtaining biological samples are also known in the art. Publications such as Ramzy, Ibrahim Clinical Cytopathology and Aspiration Biopsy 2001, which is herein incorporated by reference in its entirety, describes general methods for biopsy and cytological methods.

[0082] In some aspects of the present methods, the molecular profiling business may obtain the biological sample from a subject directly, from a medical professional, from a third party, or from a kit provided by a molecular profiling business or a third party. In some cases, the biological sample may be obtained by the molecular profiling business after the subject, a medical professional, or a third party acquires and sends the biological sample to the molecular profiling business. In some cases, the molecular profiling business may provide suitable containers, and excipients for storage and transport of the biological sample to the molecular profiling business.

[0083] In some aspects of the methods described herein, a medical professional need not be involved in the initial diagnosis or sample acquisition. An individual may alternatively obtain a sample through the use of an over the counter (OTC) kit. An OTC kit may contain a means for obtaining said sample as described herein, a means for storing said sample for inspection, and instructions for proper use of the kit. In some cases, molecular profiling services are included in the price for purchase of the kit. In other cases, the molecular profiling services are billed separately. A sample suitable for use by the molecular profiling business may be any material containing tissues, cells, nucleic acids, genes, gene fragments, expression products,300224353.1 - 28 -gene expression products, or gene expression product fragments of an individual to be tested. Methods for determining sample suitability and / or adequacy are provided.

[0084] In some aspects, the subject may be referred to a specialist such as an oncologist, surgeon, or endocrinologist. The specialist may likewise obtain a biological sample for testing or refer the individual to a testing center or laboratory for submission of the biological sample. In some cases the medical professional may refer the subject to a testing center or laboratory for submission of the biological sample. In other cases, the subject may provide the sample. In some cases, a molecular profiling business may obtain the sample.IV. Metabolite and Protein Biomarkers

[0085] Aspects herein relate to the measurement of metabolites and / or proteins. Any of the metabolites and / or proteins disclosed herein may be measured in certain aspects herein.

[0086] In certain aspects, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, or 11 biomarkers (or any range derivable therein) biomarkers of Table 1 are measured.Table 1: Metabolites for Aspects Herein

[0087] In certain aspects, 1, 2, 3, 4, 5, 6, 7, or 8 biomarkers (or any range derivable therein) of Table 2 are measured.Table 2: Biomarkers for Aspects Herein300224353.1 - 29 -

[0088] In certain aspects, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, or 11 biomarkers (or any range derivable therein) biomarkers of Table 3 are measured.Table 3: Metabolites for Aspects Herein

[0089] In certain aspects, 1, 2, 3, 4, 5, 6, 7, 8, or 9 biomarkers (or any range derivable therein) biomarkers of Table 4 are measured.Table 4: Biomarkers for Aspects Herein300224353.1 - 30 -

[0090] In certain aspects, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, or 15 biomarkers (or any range derivable therein) biomarkers of Table 5 are measured.Table 5: Biomarkers for Aspects Herein

[0091] In certain aspects, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, or 15 biomarkers (or any range derivable therein) biomarkers of Table 5 are measured.Table 6: Biomarkers for Aspects HereinV. Administration of Therapeutic Compositions

[0092] The therapy provided herein may comprise administration of a combination of therapeutic agents, such as a first therapy and a second therapy. The therapies may be administered in any suitable manner known in the art. For example, the first and second treatment may be administered sequentially (at different times) or concurrently (at the same300224353.1 - 31 -time). In some aspects, the first and second treatments are administered in a separate composition. In some aspects, the first and second treatments are in the same composition.

[0093] Aspects of the disclosure relate to compositions and methods comprising therapeutic compositions. The different therapies may be administered in one composition or in more than one composition, such as 2 compositions, 3 compositions, or 4 compositions. Various combinations of the agents may be employed.

[0094] The therapeutic agents of the disclosure may be administered by the same route of administration or by different routes of administration. In some aspects, the therapy is administered intravenously, intramuscularly, subcutaneously, topically, orally, transdermally, intraperitoneally, intraorbitally, by implantation, by inhalation, intrathecally, intraventricularly, or intranasally. In some aspects, the antibiotic is administered intravenously, intramuscularly, subcutaneously, topically, orally, transdermally, intraperitoneally, intraorbitally, by implantation, by inhalation, intrathecally, intraventricularly, or intranasally. The appropriate dosage may be determined based on the type of disease to be treated, severity and course of the disease, the clinical condition of the individual, the individual's clinical history and response to the treatment, and the discretion of the attending physician.

[0095] The treatments may include various “unit doses.” Unit dose is defined as containing a predetermined-quantity of the therapeutic composition. The quantity to be administered, and the particular route and formulation, is within the skill of determination of those in the clinical arts. A unit dose need not be administered as a single injection but may comprise continuous infusion over a set period of time. In some aspects, a unit dose comprises a single administrable dose.

[0096] Precise amounts of the therapeutic composition also depend on the judgment of the practitioner and are peculiar to each individual. Factors affecting dose include physical and clinical state of the patient, the route of administration, the intended goal of treatment (alleviation of symptoms versus cure) and the potency, stability and toxicity of the particular therapeutic substance or other therapies a subject may be undergoing.VI. Pharmaceutical Compositions

[0097] In certain aspects, the compositions or agents for use in the methods, such as therapeutic agents or biomarker modulators, are suitably contained in a pharmaceutically acceptable carrier. The carrier is non-toxic, biocompatible and is selected so as not to detrimentally affect the biological activity of the agent. The agents in some aspects of the disclosure may be formulated into preparations for local delivery (i.e. to a specific location of the body, such as skeletal muscle or other tissue) or systemic delivery, in solid, semi-solid, gel,300224353.1 - 32 -liquid or gaseous forms such as tablets, capsules, powders, granules, ointments, solutions, depositories, inhalants and injections allowing for oral, parenteral or surgical administration. Certain aspects of the disclosure also contemplate local administration of the compositions by coating medical devices and the like.

[0098] Suitable carriers for parenteral delivery via injectable, infusion or irrigation and topical delivery include distilled water, physiological phosphate-buffered saline, normal or lactated Ringer's solutions, dextrose solution, Hank's solution, or propanediol. In addition, sterile, fixed oils may be employed as a solvent or suspending medium. For this purpose any biocompatible oil may be employed including synthetic mono- or diglycerides. In addition, fatty acids such as oleic acid find use in the preparation of injectables. The carrier and agent may be compounded as a liquid, suspension, polymerizable or non-polymerizable gel, paste or salve.

[0099] The carrier may also comprise a delivery vehicle to sustain (i.e., extend, delay or regulate) the delivery of the agent(s) or to enhance the delivery, uptake, stability or pharmacokinetics of the therapeutic agent(s). Such a delivery vehicle may include, by way of non-limiting examples, microparticles, microspheres, nanospheres or nanoparticles composed of proteins, liposomes, carbohydrates, synthetic organic compounds, inorganic compounds, polymeric or copolymeric hydrogels and polymeric micelles.

[0100] In certain aspects, the actual dosage amount of a composition administered to a patient or subject can be determined by physical and physiological factors such as body weight, severity of condition, the type of disease being treated, previous or concurrent therapeutic interventions, idiopathy of the patient and on the route of administration. The practitioner responsible for administration will, in any event, determine the concentration of active ingredient(s) in a composition and appropriate dose(s) for the individual subject.

[0101] Solutions of pharmaceutical compositions can be prepared in water suitably mixed with a surfactant, such as hydroxypropylcellulose. Dispersions also can be prepared in glycerol, liquid polyethylene glycols, mixtures thereof and in oils. Under ordinary conditions of storage and use, these preparations contain a preservative to prevent the growth of microorganisms.

[0102] In certain aspects, the pharmaceutical compositions are advantageously administered in the form of injectable compositions either as liquid solutions or suspensions; solid forms suitable or solution in, or suspension in, liquid prior to injection may also be prepared. These preparations also may be emulsified. A typical composition for such purpose comprises a pharmaceutically acceptable carrier. For instance, the composition may contain 10 mg or less, 25 mg, 50 mg or up to about 100 mg of human serum albumin per milliliter of300224353.1 - 33 -phosphate buffered saline. Other pharmaceutically acceptable carriers include aqueous solutions, non-toxic excipients, including salts, preservatives, buffers and the like.

[0103] Examples of non-aqueous solvents are propylene glycol, polyethylene glycol, vegetable oil and injectable organic esters such as ethyloleate. Aqueous carriers include water, alcoholic / aqueous solutions, saline solutions, parenteral vehicles such as sodium chloride, Ringer's dextrose, etc. Intravenous vehicles include fluid and nutrient replenishers. Preservatives include antimicrobial agents, antgifungal agents, anti-oxidants, chelating agents and inert gases. The pH and exact concentration of the various components the pharmaceutical composition are adjusted according to well-known parameters.

[0104] Additional formulations are suitable for oral administration. Oral formulations include such typical excipients as, for example, pharmaceutical grades of mannitol, lactose, starch, magnesium stearate, sodium saccharine, cellulose, magnesium carbonate and the like. The compositions take the form of solutions, suspensions, tablets, pills, capsules, sustained release formulations or powders.

[0105] In further aspects, the pharmaceutical compositions may include classic pharmaceutical preparations. Administration of pharmaceutical compositions according to certain aspects may be via any common route so long as the target tissue is available via that route. This may include oral, nasal, buccal, rectal, vaginal or topical. Alternatively, administration may be by orthotopic, intradermal, subcutaneous, intramuscular, intraperitoneal or intravenous injection. Such compositions would normally be administered as pharmaceutically acceptable compositions that include physiologically acceptable carriers, buffers or other excipients. For treatment of conditions of the lungs, aerosol delivery can be used. Volume of the aerosol is between about 0.01 ml and 0.5 ml.

[0106] An effective amount of the pharmaceutical composition is determined based on the intended goal. The term “unit dose” or “dosage” refers to physically discrete units suitable for use in a subject, each unit containing a predetermined-quantity of the pharmaceutical composition calculated to produce the desired responses discussed above in association with its administration, i.e., the appropriate route and treatment regimen. The quantity to be administered, both according to number of treatments and unit dose, depends on the protection or effect desired.

[0107] Precise amounts of the pharmaceutical composition also depend on the judgment of the practitioner and are peculiar to each individual. Factors affecting the dose include the physical and clinical state of the patient, the route of administration, the intended goal of300224353.1 - 34 -treatment (e.g., alleviation of symptoms versus cure) and the potency, stability and toxicity of the particular therapeutic substance.VII. Kits

[0108] Certain aspects of the present invention also concern kits containing compositions of the invention or compositions to implement methods of the invention. In some aspects, kits can be used to evaluate one or more biomarkers. In certain aspects, a kit contains, contains at least or contains at most 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 100, 500, 1,000 or more probes, primers or primer sets, synthetic molecules, antibodies, or inhibitors, or any value or range and combination derivable therein. In some aspects, there are kits for evaluating biomarker activity or level in a cell.

[0109] Kits may comprise components, which may be individually packaged or placed in a container, such as a tube, bottle, vial, syringe, or other suitable container means.

[0110] Individual components may also be provided in a kit in concentrated amounts; in some aspects, a component is provided individually in the same concentration as it would be in a solution with other components. Concentrations of components may be provided as 1x, 2x, 5x, 10x, or 20x or more.

[0111] Kits for using probes, antibodies, synthetic nucleic acids, nonsynthetic nucleic acids, and / or inhibitors of the disclosure for prognostic or diagnostic applications are included as part of the disclosure. Specifically contemplated are any such molecules corresponding to any biomarker identified herein, which includes antibodies that bind to such biomarkers as well as nucleic acid primers / primer sets and probes that are identical to or complementary to all or part of a biomarker, which may include noncoding sequences of the biomarker, as well as coding sequences of the biomarker.

[0112] In certain aspects, negative and / or positive control nucleic acids, antibodies, probes, and inhibitors are included in some kit aspects. In addition, a kit may include a sample that is a negative or positive control for methylation of one or more biomarkers.

[0113] It is contemplated that any method or composition described herein can be implemented with respect to any other method or composition described herein and that different aspects may be combined. The claims originally filed are contemplated to cover claims that are multiply dependent on any filed claim or combination of filed claims.VIII. Examples

[0114] The following examples are included to demonstrate preferred embodiments of the invention. It should be appreciated by those of skill in the art that the techniques disclosed in300224353.1 - 35 -the examples which follow represent techniques discovered by the inventor to function well in the practice of the invention, and thus can be considered to constitute preferred modes for its practice. However, those of skill in the art should, in light of the present disclosure, appreciate that many changes can be made in the specific embodiments which are disclosed and still obtain a like or similar result without departing from the spirit and scope of the invention.Example 1 - Propensity score-matched CASH & non-CASH subjects were not significantly different in age, sex, sporadic / familial genotype, brainstem lesion locationExample 2 - Materials and methods for aspects disclosed hereinDifferentially expressed metabolites in discovery diagnostic & prognostic CASH cohorts

[0115] Two propensity-matched cohorts of patients with Cerebral Cavernous Malformation (CCM) patients were enrolled in this prospective study. A diagnostic cohort of was constituted of 20 CCM patients who experienced symptomatic hemorrhage during the year prior to blood collection and to 20 propensity-matched CCM patients without clinical events during the prior year. The patients were propensity-matched for (1) age at enrollment, (2) gender, (3)300224353.1 - 36 -familial / sporadic-CM, (4) CCM location (non-brainstem / brainstem), and (5) epilepsy (>1 seizure in the prior year, with / without medications).

[0116] The prognostic cohort included 15 CCM patients that developed clinical events during the year following blood collection, propensity -matched to 15 CCM patients without clinical events during the following year. The patients were propensity-matched for (1) age at enrollment, (2) gender, (3) familial / sporadic-CCM, (4) CCM location (non-brainstem / brainstem), (5) epilepsy (>1 seizure in the prior year, with / without medications), and (6) prior SH in the year before enrolment. The diagnosis of CCM and associated clinical events were confirmed by an experienced senior neurosurgeon (IAA) through clinical 3T MRI scans.Sample Extraction and Processing for LC-MS Assays

[0117] Metabolites were extracted using extraction solvent (100% methanol spiked with heavy labeled internal standards and stored at -80°C) added to each plasma sample at a ratio of 50pL of plasma per 400pL of extraction solvent in microcentrifuge tubes. Samples were then vortexed for 30 seconds and centrifuged at -10°C, 20,000xg for 15min and the supernatant was subsequently used for metabolomic analysis. Seventy-five-pL of metabolite extract was added to prelabeled microcentrifuge tubes and dried down completely under vacuum using a Genevac EZ-2 Elite (Genevac Ltd., Ipswich, UK)https: / / www.bionity.com / en / companies / 12303 / genevac-ltd.html. Samples were resuspended in 200 pL of 5% Acetonitrile, and then added to a thermomixer (Eppendorf, Hamburg, Germany) to resuspend analytes at 4 °C, 1000 rpm for 15 min with an infinite hold at 4 °C. Samples were finally centrifuged at 4°C, 20,000 x g for 15 min to remove insoluble debris and lOOpL of supernatant was transferred to a prelabeled MS vial.

[0118] Samples were analyzed on Thermo Fisher ultra-performance liquid chromatography system coupled to an Orbitrap IQ-X mass spectrometer, operating in positive mode. Five-pL of sample was injected onto an Cortecs© UPLC T3 Column (1.2 pm, 2.1 x 100 mm) fitted with Cortecs© UPLC T3 guard at 30 °C. The mobile phase A was water with 5% Acetonitrile and 0.1% Formic Acid and mobile phase B was 100% Acetonitrile with 0.1% Formic Acid. Gradient elution started with 0% B with a flow rate of 0.40 mL / min for 0.2 min and linearly increased to 100% B over 10 min and these conditions were held constant for 1.0 min. Finally, re-equilibration at 0% B was performed for 1.5 min. The electrospray ionization conditions were set with the spray voltage at 3.4 kV, vaporizer temp at 400°C, and detection window set to 120-1500 m / z. Internal mass calibration was performed with continuous infusion of Easy-IC© at the MS1level. A Deep Scan AcquireX data acquisition workflow was performed in which precursor selection for MS2scans was set to 120-1500 m / z with Targeted mass list300224353.1 - 37 -created by repeated injection of a pooled QC sample and a targeted exclusion list created by injection of a method blank. The isolation window was 0.7 m / z with no offset with stepped collision energies of 20, 40, 60, and 80 V combined into one MS2spectrum.

[0119] Raw data were processed in Mzmine 2.53 to produce a list of features (an m / z at a particular retention time) with the normalized peak area of that feature found in each sample. Noise was filtered out for mass detection by setting intensity cutoffs. Only signals above 4E4 in the MS1level and 1E4 in the MS2level were included, m / z and retention time tolerances along with peak shape criteria were set to extract individual peaks. The IQX has high mass and retention time accuracy, so tolerances are relatively strict (0.005 m / z and 0.10 min). Features within m / z and retention time tolerances that likely represent isotopologues were grouped together rather than considered two different features. Raw peak areas were normalized using a weighted contribution of all selected internal standards, weighted by distance. The distance of the standard peak to the peak being normalized was calculated as distance = (MZ difference) + (RT difference). All features with associated MS2scans were used to create a subset of features to be analyzed in GNPS. All MS2scans for a single feature were averaged to create a consensus spectrum for each feature. The cut-off for removing features based on Polled QC sample is to remove any feature with a %Coeffi cient of Variation (%CV) greater than 20%.Subclass networking analyses

[0120] GNPS is used to analyze the collected MS2 data. GNPS (https: / / gnps.ucsd.edu) is an open-access platform that contains libraries of tandem mass spectrometry data.1Data analyzed in GNPS is not public until uploaded by the user to MassIVE for community use. The platform allows users to search their data for fragmentation patterns that match or are analogous to known library fragmentation patterns, and simultaneously create networks of related fragmentation patterns in their dataset. Different workflows are available in GNPS: Feature Networking uses data that has been curated into feature lists and the classical Molecular Networking uses unprocessed, raw data files. However, the latter workflow can include data that gets filtered out during feature list curation and MS2 spectra are combined into consensus spectra in a different manner than during feature list curation with can result in idiosyncrasies between the two methods.

[0121] The mass spectrometry data was first processed with MZMINE22and the results were exported to GNPS for FBMN analysis. A molecular network was created with the Feature-Based Molecular Networking (FBMN) workflow3on GNPS (https: / / gnps.ucsd.edu).4The data was filtered by removing all MS / MS fragment ions within + / - 17 Da of the precursor m / z. MS / MS spectra were window filtered by choosing only the top 6 fragment ions in the + / -300224353.1 - 38 -50 Da window throughout the spectrum. The precursor ion mass tolerance was set to 0.02 Da and the MS / MS fragment ion tolerance to 0.02 Da. A molecular network was then created where edges were filtered to have a cosine score above 0.7 and more than 4 matched peaks. Further, edges between two nodes were kept in the network if and only if each of the nodes appeared in each other’s respective top 10 most similar nodes.

[0122] Finally, the maximum size of a molecular family was set to 100, and the lowest scoring edges were removed from molecular families until the molecular family size was below this threshold. The analogue search mode was used by searching against MS / MS spectra with a maximum difference of 100.0 in the precursor ion value. The library spectra were filtered in the same manner as the input data. All matches kept between network spectra and library spectra were required to have a score above 0.7 and at least 4 matched peaks. The DEREPLICATOR5was used to annotate MS / MS spectra. The molecular networks were visualized using Cytoscape software (http: / / www.cytoscape.org / ).6Monitoring BiomarkerStudy design and population

[0123] This is a case-control prospective study that includes 46 subjects enrolled between 29 / 05 / 2020 and 27 / 09 / 2021 within the CASH Trial Readiness project7(clinicaltrials.gov; NCT03652181) at the University of Chicago Medicine (UCM). Among the 46 subjects enrolled, 32 contributed for one yearly longitudinal follow-up while 14 for two yearly longitudinal follow-ups were included.

[0124] Patients (1) 18 years or older, (2) harboring one or multiple CCM in their brain and (3) experienced a symptomatic hemorrhage (SH) within the prior year in a lesion without prior or planned treatment (resection or irradiation) were enrolled. A symptomatic hemorrhage (SH) is defined by demonstrated new lesional bleeding or hemorrhagic growth on imaging and attributable new or worsening symptoms, as per current clinical guidelines.8Exclusion criteria included (1) prior brain irradiation, (2) contraindication for contrast agent administration or (3) unwillingness to undergo research MRI studies, (4) pregnancy in the past 6 months or (5) breastfeeding, (6) inability to verify SH with clinical and imaging review, and / or (7) any reason for unlikeliness to return for follow-up visits. This clinical study was approved by the central institutional review board at (UCM) (IRB 20-0518). The subjects’ consent was obtained according to the Declaration of Helsinki.Definition of the Imaging Biomarkers Outcomes and Circulating Compound Variables

[0125] Patients enrolled underwent quantitative susceptibility mapping (QSM) that measures the spatial distribution of iron deposition as previously published.9, 10Dynamic300224353.1 - 39 -contrast-enhanced quantitative permeability (DCEQP) assesses the vascular permeability as reported previously.9, 11

[0126] Categorical Change for Imaging Biomarkers (i.e., QSM and / or DCEQP). A categorical binary variable (Yes / No) was defined as a yearly change in mean lesional QSM>6% or DCEQP > 40% biomarkers as previously reported.

[0127] Relative Percent Change. A relative percent change (RpC) in QSM or DCEQP lesional mean value and in plasma levels of protein or metabolite was defined as:100 [1]X: Value of the imaging biomarkers (i.e., QSM or DCEQP) or the plasma levels of a specific circulating compound (i.e., protein or metabolite) at to the 1stpoint in epoch, and ti the 2ndpoint in epoch.

[0128] Absolute Change. An absolute change in the lesional mean value of QSM or DCEQP and in plasma levels of protein or metabolite was defined as:Absolute change of Biomarker X = Biomarker Xtl— Biomarker Xt0) [2] X: ti and to as defined in equation [1],Statistical Analyses

[0129] Correlation Analyses. Correlation analyses between mean lesional QSM or DCEQP values and the plasma levels of protein or metabolite were first conducted using Pearson correlation. A strong correlation (\p | > 0.5) between the imaging biomarkers (i.e., QSM and / or DCEQP) and the circulating compounds (i.e., proteins or metabolites) may suggest a surrogate.

[0130] Univariate Analyses. Univariate analyses first assessed if a (1) relative change or (2) absolute change in plasma levels of circulating compounds (i.e., protein or metabolite estimated using equation [1] and [2], respectively) can reflect a yearly change in mean lesional QSM>6% (Yes / No) or DCEQP > 40% (Yes / No). These analyses were performed using a logistic regression. Further univariate analyses to test if a (1) relative or (2) absolute change in plasma levels of circulating compounds (i.e., protein or metabolite) reflect a (1) relative or (2) absolute in mean lesional QSM or DCEQP. These analyses were performed using a linear regression with a false discovery rate (FDR) correction.

[0131] Multivariate Analyses. Multivariate analyses assessed if a (1) categorical, (2) relative or (3) absolute change in mean lesional QSM or DCEQP can reflect a change (i.e., relative or absolute) in plasma levels of a circulating compounds (i.e., protein or metabolite). These multivariate analyses were performed using only the plasma levels of (1) the proteins (2) the metabolites and then (3) a combination of both (i.e., protein and metabolite). All possible combinations of at least two proteins and / or metabolites were assessed using leave-one-out300224353.1 - 40 -cross validation (LOOCV). Each model was evaluated using three metrics (e.g., the smaller the better): (1) sum of squared error (SSE) from LOOCV, which assess the residual between the predicted results to the observation, (2) Akaike information criterion (AIC), which penalizes the inclusion of too co-variates, (3) error rate (ER), which is the percent of wrong signs between the categorical, relative or absolute change observed in the imaging biomarkers (i.e., QSM / DCEQP) and the prediction made by the multivariate model.

[0132] All statistical analyses were performed on R statistical framework (v3.4, R Foundation for Statistical Computing, https: / / www.r-project.org / ).12Example 3 - Plasma Metabolites as Biomarker for CASH

[0133] As shown in the Appendix, the inventors have identified sets of metabolites useful for diagnosing, prognosing, and monitoring the progression of CASH.

[0134] Background. An increase in mean lesional iron content (>6%), measurement by quantitative susceptibility mapping (QSM), and / or vascular permeability (>40%), assessed by dynamic contrast enhanced quantitative perfusion (DCEQP), have been associated with new symptomatic hemorrhage (SH) in cerebral cavernous malformations. These 2 imaging sequences are being assessed as surrogate outcomes in clinical trials of novel pharmacotherapy aimed at decreasing rebleeding in CCM. Of interest, plasma proteins and metabolites have been associated with hemorrhagic activity of CCMs and have been mechanistically linked to the permissive microbiome and lesional transcriptome of CCM. However, changes in the plasma levels of these circulating molecules have not been compared to prospective changes in mean QSM and DCEQP.

[0135] Methods. Forty-six CCM patients with SH in the prior year enrolled in the NIH multicenter Trial Readiness (U01 NS 104157) were included in the study. The plasma samples and QSM as well as DCEQP dataset from these 46 patients were simultaneously acquired at the beginning and end of 60 one-year epochs of prospective follow-up. Plasma levels of 16 proteins and 12 metabolites previously associated with CCM hemorrhage were assessed by ELISA and liquid-chromatography mass spectrometry, respectively. Changes in plasma levels (i.e., relative and absolute) in relation to an increase in mean lesional QSM>6% (Yes / No) and / or DCEQP>40% (Yes / No) during a 1-year epoch were assessed using univariate and multivariate analyses. Multi-omic combinations of plasma levels of proteins and metabolites reflecting QSM and / or DCEQP changes were selected based on the sum of squared error (SSE) from LOOCV, accuracy (sensitivity / specificity on receiver operating curves), and the biomarker’s error rate.300224353.1 - 41 -

[0136] Results. None of the proteins individually correlated with mean increase of QSM>6% or DCEQP>40% in univariate analyses. Only the relative change in plasma levels of acetyl. L. carnitine was correlated with an increase in mean lesional QSM>6% (91.5% and 100% sensitivity / specificity, / ?=0.01). Further multivariate analyses showed that a multi-omic model combining the relative changes in plasma levels of three proteins (i.e., roundabout guidance receptor-4, cluster of differentiation 14, thrombomodulin) and 1 metabolite (i.e., acetyl. L. carnitine) reflected a mean increase in QSM>6% (97.2% and 100% specificity / sensitivity). A combination of relative changes in plasma levels of endoglin and 3 metabolites including pipecolic and arachidonic acids as well as hypoxanthine correlated with an increase in mean DCEQP >40% (99.6% specificity and 100% sensitivity).

[0137] Conclusion. Multi-omic combination of plasma changes of circulating proteins and metabolites reflect with great accuracy the changes in lesional iron content and permeability during annual follow-up of CCMs with recent SH. Results have mechanistic implications based on functional pathways correlated with these molecules. They are a proof of concept that blood tests could replace more complex and costly imaging biomarkers in monitoring of CCM hemorrhage, and as secondary outcomes in clinical trials.Example 4 - Circulating Molecules Reflect Imaging Biomarkers of Hemorrhage in Cerebral Cavernous Malformations

[0138] Introduction

[0139] Cerebral cavernous malformations (CCMs), also known as cavernous angiomas, are abnormal blood filled capillaries prone to hemorrhage, with a worldwide prevalence ranging from 0.1% to 0.5% The clinical course of CCM disease remains highly variable.1, 2Patients can experience an aggressive clinical course with earlier clinical symptom onset, higher lesion burden, and more frequent symptomatic hemorrhage.3Although the annual risk of a first symptomatic hemorrhage (SH) is estimated below 1%, the risk of recurrent bleeding after a SH is ten-fold higher with additional cumulative disability as compared to quiescent lesions.1, 2It remains unclear what predisposes these patients to this increased risk,1but ongoing inflammation and angiogenesis after a SH may contribute to this process. Additionally, it has been recently reported that subclinical bleed, identified as asymptomatic change (AC) on imaging, also increases the likelihood of future SH.4There is therefore a need for accurate biomarkers to distinguish high risk cases, selecting them for invasive or novel interventions, and to track disease status, progression, and response to therapies.5, 6300224353.1 - 42 -

[0140] Quantitative susceptibility mapping (QSM) and dynamic contrast enhanced quantitative perfusion (DCEQP) have been respectively used on magnetic resonance imaging (MRI) to assess iron content and vascular permeability in CCM lesions7, 8An increase in mean lesional QSM of >6% (sensitivity, 82.3%; specificity, 88.89%) and an increase of >40% (sensitivity, 78.72%; specificity, 88.89%) in mean lesional DCEQP were associated with new SH during annual follow-up of previously stable CCMs.9More recently, these were validated as monitoring biomarkers of new bleeding in CCMs with recent SH, also referred to as cavernous angioma with symptomatic hemorrhage (CASH),10with perfect specificity and greater sensitivity to CCM bleeding than clinical manifestations. Even more, QSM has been accepted by the United States Food and Drug Administration as a biologically plausible biomarker for assesing drug effects on CCM hemorrhage (Biomarker Qualification Acceptance DDT-BMQ-000127). They are being used as surrogate outcomes for effect of atorvastatin on CCM bleeding in an ongoing prospective randomized double blinded clinical trial (clinicaltrials.gov NCT02603328).11

[0141] Several dysregulated pathways related to anticoagulant domains,6inflammatory mechanisms, blood-brain barrier permeability, as well as endothelial tight junction stability have been associated with CCM mechanisms, including MAPK / MEKK3 / ERK3, PI3K-Akt, and Notch.12Of interest, an increase in lipopolysaccharide (LPS)-producing Gram-negative bacteria also stimulates MAPK / MEKK3 / ERK3 activity via toll-like receptor (TLR) 4 in brain microvascular endothelium, suggesting a role of the gut-brain axis in this disease.6Furthermore, a leaky gut epithelium linked with a permissive gut microbiome has been associated with CCM genesis, and microbiome differences were shown in SH cases.13Several plasma proteins linked to the above mechanisms are therefore being probed as candidate biomarkers of CCM hemorrhagic activity.5However, circulating plasma biomarkers have not yet been assessed as monitoring biomarkers of lesional bleeding, nor correlated with lesional QSM and DCEQP. Integrating multi-omic approaches by correlating changes in circulating proteins and metabolites with prospective changes in mean lesional QSM and DCEQP would cross validate the respective biomarkers and potentially enhance the understanding of CCM biology.12Establishing this correlation would provide strong evidence that less invasive blood tests could potentially replace or complement more complex and costly imaging methods.

[0142] The inventors hypothesize that CCM lesion QSM and DCEQP changes correlate with changes in circulating biomarker levels during prospective monitoring after SH.

[0143] Materials and Methods

[0144] Study design and population300224353.1 - 43 -

[0145] This prospective cohort study includes 46 subjects enrolled between 29 / 05 / 2020 and 27 / 09 / 2021 within the CASH Trial Readiness project (clinicaltrials.gov NCT03652181),9at the University of Chicago Medicine. Among the 46 subjects enrolled, 32 contributed one annual longitudinal follow-up, and 14 contributed two annual longitudinal follow-ups with QSM and DCEQP imaging, with each subject also contributing a peripheral blood sample at the same time as baseline or annual follow-up imaging (Supplementary Table 1).

[0146] Supplementary Table 1. Subjects’ Demographics and Cerebral Cavernous Malformation Features*Location of the cavernous angioma with symptomatic hemorrhage that had bled within the prior year and considered for patient enrollment.

[0147] Patients (1) 18 years or older, (2) harboring one or multiple CCM in their brain and (3) experienced a symptomatic hemorrhage (SH) within the prior year in a lesion with no prior or planned treatment (i.e., resection or irradiation) were enrolled. A symptomatic hemorrhage (SH) is defined by demonstrated new lesional bleeding or hemorrhagic growth on imaging and attributable new or worsening symptoms, as per current clinical guidelines.2Exclusion criteria included (1) prior brain irradiation, (2) contraindication for contrast agent administration or (3)300224353.1 - 44 -unwillingness to undergo research MRI studies, (4) pregnancy in the past 6 months or (5) breastfeeding, (6) inability to verify SH with clinical and imaging review, and / or (7) any reason for unlikeliness to return for follow-up visits. This clinical study was approved by the central institutional review board at University of Chicago Medicine (IRB 20-0518). Each subject signed a consent obtained according to the Declaration of Helsinki.

[0148] QSM and DCEQP MRI acquisition and post-processing

[0149] Patients enrolled underwent QSM and DCEQP sequences in addition to their routine clinical MRI evaluation at baseline, Year-1 and -2.9QSM sequence measures the spatial distribution of iron deposition as previously published.10This MRI sequence is a single 3D, 8 echo, spoiled gradient recalled echo T2*-weighted sequence covering the whole brain. The QSM maps were post-processed using the morphology enable dipole inversion Toolbox (Medimagemmetric, New York City, New York, v1.0.464. GE.197dlc3). The mean lesional QSM of the indexed CASH lesions at baseline, Year-1- and -2 were obtained by averaging the voxel QSM values within region of interests (ROIs) encompassing the lesion across multiple images using ImageJ (Bethesda, MD, USA, v1.52d).

[0150] The DCEQP sequence assesses the vascular permeability as reported previously, and has been validated in CCMs.10The field of view covered the indexed CASH lesion.7’10The vascular permeability maps were then calculated using the Patlak's voxel-wise method modeling a two-compartment model, as previously published. Mean lesional DCEQP values were then quantified by drawing a ROI on the co-registered axial T2-weighted image showing the largest diameter of the indexed CASH lesion on a and the corresponding hemosiderin ring using ImageJ (Bethesda, MD, USA, 1.52d).7The mean lesional permeability values were then calculated by averaging the voxel values within the ROI. The post-processing pipeline was implemented in MATLAB (MathWorks, Natick, USA, v. R2022b). More details on imaging biomarker acquisition and processing are presented in Example 5. FIGS. 8A-8H presents absolute and percent change in mean lesional change in QSM and DCEQP in relation to prospective SH in recent trial readiness project.10Cases with recurrent SH2or AC representing new subclinical bleeding or growth4were noted during the same annual epochs as QSM and DCEQP change.

[0151] Circulating metabolites

[0152] An untargeted metabolomic analysis was first performed in independent discovery cohorts of CCM patients with and without SH, propensity-matched for age, sex, sporadic / solitary or familial / multifocal, brainstem location and seizure in the prior year. The300224353.1 - 45 -details of diagnostic and prognostic discovery cohorts and the process of identification of candidate metabolites are presented in Example 5.

[0153] The inventors identified 12 metabolites, with significantly different levels in independent cohorts with and without CCM bleeding (p<0.05 false discovery rate [FDR]) corrected), with known or probable structure, excluding medications subject is taking (Supplementary Table 2). Authentic standards were then purchased and matched to the selected twelve metabolites to confirm IDs. The identification confidence for each differentially expressed plasma metabolites was categorized by levels (1-5) as previously described (Supplementary Table 3).

[0154] Supplementary Table 2. Twelve Metabolites* with Diagnostic and Prognostic Associations with Cerebral Cavernous Malformation Symptomatic Hemorrhage (SH) Identified in a Propensity Matched Discovery Cohort using Untargeted Analyses*Metabolites with identified or probable structure, excluding medications subject is taking, with levels significantly different in discovery cohorts with and without SH (p<0.05 FDR corrected)

[0155] Supplementary Table 3. Data Quality Control Measurements for Authentic Standards Prepared24300224353.1 - 46 -

[0156] Finally, a targeted LC-MS analysis of the 12 metabolites was performed using 15 μl of plasma from baseline, year -1 and -2 of the enrolled 46 patients enrolled in this study. Metabolites were reported in normalized peak area, with the heavy labeled internal standards spiked into that sample used as the normalization factor. A subclass networking analyses was performed using the MS2data was analyzed in GNPS (https: / / gnps.ucsd.edu / ProteoSAFe / static / gnps-splash.jsp).

[0157] Circulating proteins

[0158] Sixteen circulating proteins were selected, based on prior associations with CCM bleeding.5, 14, 15Plasma aliquots were processed following a previously published protocol.5, 14Plasma proteins levels (ng / ml) of vascular endothelial growth factor-A (VEGF), interleukin (IL)-10, endoglin, angiopoietin-1 and -2, thrombospondin- 1 (TSP1) and -2 (TSP2), thrombomodulin (TM), cluster of differentiation- 14 (CD 14), c-reactive protein, roundabout guidance receptor 4 (ROBO4), and lipopolysaccharide binding protein (LBP) were assayed using commercial enzyme-linked immunosorbent assay kits (R& D systems, Minneapolis, Minnesota, USA). Toll-like receptor 4 (TLR4) plasma levels were assessed using (RayBio, Peachtree Corners, GA, USA). In addition, the plasma levels of IL17A, IL16 and ILip were measured using a mesoscale detection multiplex assay system (Meso Scale Diagnostics, Rockville, MD, USA). Details of these assays in CCM patients have been published previously5, 14, 15and are further described in the Example 5.

[0159] Definition of the imaging biomarkers and circulating molecules variables

[0160] A relative percent change in QSM or DCEQP lesional mean value and in plasma levels of protein or metabolite was defined as:100 [1]300224353.1 - 47 -

[0161] X: Value of the biomarkers (i.e., imaging or circulating) at the 1sttime-point (to), and at the 2nd(ti) time-point.

[0162] An absolute change in the lesional mean value of QSM or DCEQP and in plasma levels of protein or metabolite was defined as:Absolute change of Biomarker X = Biomarker Xtl— Biomarker Xt0) [2]

[0163] X: ti and to as defined in equation (1).

[0164] A categorical binary variable (Yes / No) was defined as an annual change in mean lesional QSM>6% or DCEQP > 40% biomarkers as previously reported (FIGS. 8A-8H).9’10

[0165] Statistical analyses

[0166] Correlations between mean lesional QSM or DCEQP values and the plasma levels of protein or metabolite were first conducted using Pearson correlation. A strong correlation (| | > 0.5) between the imaging biomarkers (i.e., QSM and / or DCEQP) and the circulating compounds (i.e., proteins or metabolites) was considered to suggest a surrogate.

[0167] Univariate analyses first assessed if a (1) relative change or (2) absolute change in plasma levels of circulating compounds (i.e., protein or metabolite estimated using equation (1) and (2), respectively) can reflect a yearly change in mean lesional QSM>6% (Yes / No) or DCEQP > 40% (Yes / No). These analyses were performed using a logistic regression. Further univariate analyses to test if a (1) relative or (2) absolute change in plasma levels of circulating compounds (i.e., protein or metabolite) reflect a (1) relative or (2) absolute in mean lesional QSM or DCEQP. These analyses were performed using a linear regression with a FDR correction.

[0168] Multivariate analyses assessed if a (1) categorical, (2) relative or (3) absolute change in mean lesional QSM or DCEQP can reflect a change (i.e., relative or absolute) in plasma levels of a circulating compounds (i.e., protein or metabolite). These multivariate analyses were performed using only the plasma levels of (1) the proteins (2) the metabolites and then (3) a combination of both (i.e., protein and metabolite). All possible combinations of at least two proteins and / or metabolites were assessed using leave-one-out cross validation (LOOCV). Each model was evaluated using three metrics (e.g., the smaller the better): (1) sum of squared error (SSE) from LOOCV, which assess the residual between the predicted results to the observation, (2) Akaike information criterion, which penalizes the inclusion of too co-variates, (3) error rate, which is the percent of wrong signs between the categorical, relative or absolute change observed in the imaging biomarkers (i.e., QSM / DCEQP) and the prediction made by the multivariate model.300224353.1 - 48 -

[0169] All statistical analyses were performed on R statistical framework. Multi-omic integrative analyses of metabolome, proteome, microbiome and transcriptome are described in Example 5.

[0170] Results

[0171] Correlations between blood biomarkers and imaging biomarkers at single timepoints

[0172] The preliminary univariate correlation analyses of the circulating biomarkers (i.e., proteins and metabolites) showed that only the plasma levels of angiopioetinl and hypoxanthine were correlated ( =0.58). No significant correlation were observed between the plasma levels of any of the proteins or metabolites assayed and the mean lesional QSM and / or DCEQP values at single points in time.

[0173] Change in proteins plasma levels vs. change in mean lesional QSM

[0174] Changes in individual plasma protein levels did not correlate with changes in mean lesional QSM (i.e., either absolute or relative). Correlation of combined changes in protein levels and mean lesional QSM were stronger than correlations of combined absolute changes in protein levels (FIGS. 9A-9B).

[0175] A weighted combination of relative change in three proteins (ROBO4, CD 14 and TM) levels strongly reflected with 95% sensitivity and 81% specificity (area under the curve [AUC]=95.6% [90.5%; 100%, / ?= 8.2xlO'n; cutoff value=0.33) the categorical change in mean lesional QSM, and differentiated all cases of SH and AC (FIG. 3 A):

[0176] biQSM ~ -0.03*[ROBO4(%)] -0.06*[CD14(%)] + 0.27* [TM(%)] -0.69

[0177] The combination of absolute change in levels of only two proteins reflected the categorical change in mean lesional QSM with slightly weaker accuracy (AUC =90.2% [78.8%; 100%]; / ?=1.2xl0'8; sensitivity 86.0%, specificity 85%; cutoff value= 0.67), also differentiating cases with SH and AC (FIGS. 9A-9B).

[0178] Change in metabolites plasma levels vs. change in mean lesional QSM

[0179] Only an absolute change in acetyl. L. carnitine individually correlated weakly with absolute change in mean QSM (p=0.43, r2=0.18; p=0.001') (FIG. 11). Correlations of combination of metabolite plasma levels with relative change in mean QSM were weak (FIG.12).

[0180] The combination of relative change in two metabolites demonstrated 95.0% sensitivity and 87.0% specificity to the categorical change in mean lesional QSM (4 / / C=95.9%300224353.1 - 49 -[91.5%; 100%] / 2=1.5xl0’n; cuttoff value= 0.34), and this combined weighted model differentiated all cases with SH and AC (FIG. 3B):

[0181] biQSM ~ -0.16* [acetyl. L. carnitine (%) ] -8.5xl0~4*[lysophosphatidylethanolamine (LPE) 18:0 (%)] -0.65

[0182] A different weighted combination of the absolute change in two metabolites levels reported a weaker correlation to the categorical change in mean lesional QSM (FIG. 13).

[0183] Change in proteins and metabolites plasma levels combinations vs. change in mean lesional QSM

[0184] The combination of relative change in seven proteins and one metabolite levels was strongly correlated ( =0.66, A2=0.44, / >=0.002; SSE under LOOCV= 2.89) with a relative change in mean lesional QSM with a 86.0% sensitivity and 89.0% specificity. This combined model also differentiate all cases with SH and AC (FIG. 4A):

[0185] QSM (%)~0.30*[LBP (%)] + 0.1*[VEGF (%)] - O.14*[IL1[ (%)] -0.11* [methionine (%)] + 0.04*[TM (%)] + 0.02*[auc

[0186] (%)] + 4.9xlO~4*[IL16 (%)] + 1.9x10~6* [TSP 1 (%)]- 0.01

[0187] The combination of the relative change in plasma levels of ROBO4, TM and CD14 as well as acetyl. L. carnitine was able to distinguish an increase of mean lesional QSM >6% (Yes / No) with 100.0 / 89.0% sensitivity / specificity (HUC=98.9% [97.2%; 100%], / ?=3.1xl0'13; cuttoff value= 0.33) (FIG. 4B):

[0188] biQSM- -0.03* [ROBO4 (%)] + 0.28*[TM (%)] -0.10*[CD14 (%)] -0.09* [acetyl. L. carnitine (%)] -0.71

[0189] Change in proteins plasma levels vs. change in mean lesional DCEQP

[0190] Univariate correlation analyses showed that the changes in plasma levels of any of the proteins assayed (i.e., either absolute or relative percent) did not correlate individually with change in mean lesional DCEQP (i.e., either absolute or relative). There were weak correlations of combined absolute and relative protein levels and absolute and relative annual changes in mean lesional DCEQP (FIGS. 14A-14B).

[0191] A combination of absolute changes in plasma levels of endoglin, TSP2 and IL-16 was also able to identify a changes (Yes / No) in mean lesional DCEQP >40% with 97% for both sensitivity and specificity (4 / / C=98.2%, [94.7%; 100%], / ?=8.8xl0'15; cutoff value =0.11), and differentiated all SH and AC cases (FIG. 5 A):

[0192] biDCEQP - 0.30*[Aendoglin] -0.01*[ATSP2] -0.004* [AIL-16] -0.72300224353.1 - 50 -

[0193] The combination of relative change in levels of three proteins reflecting the categorical change in mean lesional DCEQP >40%, achieved a slightly weaker combined weighted model (FIG. 15) and poorer discrimination of SH and AC cases.

[0194] Change in metabolites plasma levels vs. change in mean lesional DCEQP

[0195] Continuous changes for each metabolite plasma level did not correlate individually with changes in mean DCEQP (i.e., either continuous or categorical). There were weak correlations of combined absolute and relative metabolite levels and absolute and relative annual changes in mean lesional DCEQP (FIGS. 16A-16B).

[0196] The categorical change in mean lesional DCEQP (> 40% Yes / No) were detected using percent changes in plasma levels of piperine, arachidonic acid and hypoxanthine for both SH and AC cases and all three cases with 100% sensitivity and 93.0% specificity (AUC=99.4% [97.5%; 100%], p=1.4x10⁻15; cutoff value= 0.43) (FIG. 5B):

[0197] biDCEQP — 0.16* [piperine (%)] +0.13* [arachidonic acid (%)] 0.04* [hypoxanthine (%)] -0.61

[0198] A combination of the absolute change in plasma levels of three metabolites also achieved similar performance in reflecting the categorical change in mean lesional DCEQP (FIG. 17).

[0199] Change in proteins and metabolites plasma levels combinations vs. change in mean lesional DCEQP

[0200] The combination models of changes in proteins and metabolites levels reported weak correlations to continuous change in mean DCEQP (FIGS. 18A-18B).

[0201] A composite weighted model of percent change in plasma level of endoglin and pipecolic acid, arachidonic acid and hypoxanthine reflected a change >40% in mean lesional DCEQP in both SH and AC cases with 97.0% sensitivity and 100% specificity (AUC=99.9% [99.6%; 100%],p=4.1x10⁻17; cutoff value= 0.23) (FIG. 6):

[0202] biDCEQP — 0.13* [pipecolic acid (%)] -0.05* [endoglin (%)] -0.05* [arachidonic acid (%)] -0.03* [hypoxanthine (%)] -0.11

[0203] There was no combined weighted model derived that documented a higher accuracy to categorical change in mean DCEQP.

[0204] Biological significance of associated proteins and metabolites

[0205] Multi-omic integrative analyses showed that CD 14 has been associated with apoptosis and inflammation signaling, such as nuclear factor kappa B (NF-KB) signaling, TLR signaling, and MAPK signaling. Of interest, CD14 has also been previously shown to not only300224353.1 - 51 -promote macrophage activation, but also has been implicated in the expression profiles of several other immune cell types.16Because iron accumulation has been shown to trigger ferroptosis and modulate neuroinflammatory responses, these pathways may explain the ability of CD14 to correlate with changes in iron levels detected by QSM.17No pathways were however identified for R0B04 or TM. Multi-omic integrative analyses also identified eight enriched pathways ( / ?<0.05, FDR corrected; Supplementary Table 5) common between acetyl. L. carnitine and LPE 18:0. These common pathways were also related to apoptosis and immune response including, TNF and PI3K-Akt signaling pathways that have previously been associated with CCM pathogenesis.5

[0206] Supplementary Table 5: Acety-L-camitine and LPE 18:0 Common Enriched Pathways

[0207] 300224353.1 - 52 -

[0208] The multi-omic integrative analyses further identified pathways ( / ?<0.05, FDR corrected) related to immune and inflammatory activation, vascular permeability, cell growth and division, and apoptosis. Of interest, six of these pathways were related to TLR signaling, which has been shown to be activated by heme iron through the NF-KB pathway.18

[0209] It has been previously shown to modulate heme oxygenase 1 activity, a molecule known to correlate with iron deposition as measured by QSM.19, 20This pathological link to iron accumulation may provide further evidence to support the biological plausibility of these circulating molecules in acting as diagnostic and prognostic supplements to QSM.

[0210] The inventors note that IL16 has been associated with interleukin signaling.21The multi-omic analyses showed that fluid shear stress was enriched for other molecules in these models (Supplementary Table 7). And the pathway analyses showed that proteins and metabolites were also linked to apoptotic and inflammatory functions. Of interest, both the EGFR and HIF-1 pathways have been shown to regulate VEGF expression, and thus have downstream effects on vascular permeability.22, 23Furthermore, the RIG-I-like receptor pathway has been shown to promote endothelial cell dysfunction by activating inflammatory responses, ultimately leading to vascular impairments.24

[0211] The biological significance of circulating molecules associated with QSM and DCEQP changes is summarized in FIG. 7.

[0212] Supplementary Table 7: Common Enriched Pathways between ROBO4, Thrombomodulin, CD 14 and Acetyl-L-carnitine300224353.1 - 53 -

[0213] Discussion

[0214] Weighted combinations of changes in plasma proteins and metabolites have been correlated with great accuracy to changes in mean lesional QSM (i.e., iron deposition) and DCEQP (i.e., permeability) of CCM lesions implicated in lesional hemorrhage. These results suggest that circulating molecules may complement or replace quantitative imaging techniques in the monitoring of bleeding in CCMs. This is the first report, to the inventors’ knowledge, linking circulating molecules to actual pathobiologic changes within a hemorrhagic cerebrovascular lesion.

[0215] Iron deposition (assessed by QSM) and vascular permeability (assessed by DCEQP) reflect distinct pathophysiological processes

[0216] QSM and DCEQP capture distinct MRI signals that reflect different aspects of CCM pathobiology7, 10Similarly, circulating proteins and metabolites can indicate different pathological processes occurring within the lesions. The circulating molecules associated to a change in QSM, such as CD 14, ROBO4, and TM, are primarily related to inflammatory responses and vascular integrity.25'27In contrast, a change in DCEQP was related to a change in endoglin, arachidonic acid, hypoxanthine, and pipecolic acid, which have been ascociated to endothelial function, oxidative stress, and / or metabolic regulation.300224353.1 - 54 -

[0217] Inflammatory and vascular integrity proteins reflect changes in lesional iron content

[0218] The results herein showed that changes in plasma levels of CD14, ROBO4, and TM are correlated with changes in mean lesional QSM. These three compounds have been related to inflammatory pathways as well as vascular integrity.25-27Several studies have reported that CD14 can act as a co-receptor for the detection of bacterial LPS and activating inflammatory pathways such as NF-DB and TLR signaling.16Recently, CD 14 has been shown to drive CCM physiopathogenesis in a murine model which is consistent with CD14 as one of the specific receptors that macrophages use to facilitate the clearance of excess ironfrom a bleed.6, 28, 29It has also been reported that patient with higher lesion burden shows a polymorphism on CD14.6ROBO4, an endogenous inhibitor of VEGF signaling expressed by vascular endothelial cells, also plays a role in maintaining vascular integrity and endothelial barrier stability.27Increased circulating levels of ROBO4 has been associated with SH and AC.14Finally, TM stimulates the levels of anticoagulant activated protein C and has been shown to be increased in human CCM lesional milieu and elevated levels of TM6Taken together, these resuts suggest that the circulating compounds may reflect pro- and anti-inflammatory as well as anti -thrombotic processes occurring in-situ the lesion aiming to clear blood products, and observed by changes in iron deposition (i.e., assessed using QSM).

[0219] Energy and inflammation metabolites associated with changes in lesional iron content

[0220] Acetyl-L-camitine plays a role in fatty acid metabolism and has been implicated in neuroprotection and mitochondrial health.30Its correlation with QSM changes may be linked to its role in mitigating oxidative stress and modulating inflammatory responses, both of which are crucial in the context of iron deposition within the brain.31Acetyl-L-carnitine is known to reverse iron-induced oxidative stress in fibroblast and can act as an anti -angiogenic molecule reducing inflammation during angiogenesis.31, 32Acetyl-L-camitine also inhibits NF-DB and ICAM-1 reducing adhesion of monocytes on the endothelial cells.32Acetyl-L-camitine together with the LPE18:0 have been showed to stimulate arachidonic acid.33Further pathway analyses also identified fluid shear stress pathways, which have been linked to the modulation of heme oxygenase 1 activity, a known correlate of iron deposition measured by QSM.19, 20

[0221] Extracellular matrix (ECM) remodeling and TGF- pathway proteins are associated with changes in lesional vascular permeability

[0222] A change in DCEQP values have been associated to a change in plasma levels of endoglin, TSP2, and IL16. Endoglin is a critical regulator of angiogenesis and endothelial300224353.1 - 55 -function, particularly in the context of vascular remodeling.34It has been associated to CCM physiopathogenesis.35Endoglin is a co-receptor in the TGF-P signaling pathway, which is essential for maintaining endothelial cell function and vascular homeostasis.34Endoglin influences the balance between endothelial cell proliferation, differentiation, and migration, by modulating TGF-P signaling.36’37TSP2 impact vascular permeability by regulating the ECM and endothelial cell function.38Elevated levels of TSP2 have been associated with decreased vessel permeability, as it strengthens the ECM and tightens the endothelial barrier.38’39IL16 can also influence endothelial cell function, potentially leading to alterations in the integrity of the vascular barrier.40

[0223] Oxidative stress and eicosanoid pathway metabolites are associated with changes in lesional vascular permeability

[0224] The plasma levels of piperine, arachidonic acid, and hypoxanthine were also associated to changes in DCEQP. Plasma levels of hypoxanthine have been shown to be lower in CCM patients while higher levels of arachidonic acid were associated with hemorrhagic activity. Hypoxanthine is a marker of oxidative stress, which is closely linked to endothelial dysfunction41It regulates expression of apoptosis related proteins when reactive oxygen species are present41In addition, arachidonic acid has been recently linked to the regulation of vascular permeability in CCM through VEGF.42In addition, Piperine has shown to reduce angiogenesis in in vitro model and modulate inflammatory pathways inhibiting LPS induced inflammatory responses43,44This is particularly interesting since LPS is known to dramatically increase lesion genesis through TLR4-CD14-MEKK-KLF2 / 4 mechanism.6Example 5 - Supplementary Aspects for Example 4

[0225] Participants

[0226] This is not a matched study. The inventors aimed compare imaging and plasma data collected at the same time points during prospective follow-up of the same cohort of subjects who met the articulated inclusion and exclusion criteria. All patients in the cohort study contributed paired imaging and plasma data and there was no missing data in the studied cohort. Imaging and plasma data acquisition methodologies are articulated. The inventors pre-specified the occurrence of symptomatic hemorrhage (SH), subclinical bleed on MRI (asymptomatic change, AC) and the occurrence of >6% increase in QSM or >40% increase in DCEQP as index events during follow-up. These are clearly defined.

[0227] Sample size

[0228] Based on the frequency of threshold QSM or DCEQP increases during 1-year epochs in the Trial Readiness project1, and observed changes in the pilot diagnostic and prognostic300224353.1 - 56 -plasma biomarkers during 1-year epochs in pilot studies2, 50 paired observations would allow at least 90% power for detecting significant correlation between those changes (Pearson r=0.7; P=0.002).

[0229] Quantitative susceptibility mapping (QSM) and dynamic contrast enhanced quantitative perfusion (DCEQP) acquisition and processing

[0230] Imaging sequences were acquired with a 3T magnetic resonance imaging (MRI) system (Achieva; Philips Healthcare, Best, the Netherlands). QSM data were acquired with an eight channel phased array head coil, a 3D, T2*-weighted, multiecho, spoiled gradient-echo sequence with common parameters, including 8 TEs with uniform spacing, a flip angle of 15°, and a parallel acceleration factor of 2. System-specific parameters were applied as follows: TE= [5.6;51 ms]; FOV =224 mm; acquisition matrix=224 x 224; slab-encoding thickness= 1 mm; TR=66 ms.3’5

[0231] QSM images were reconstructed using the morphology-enabled dipole inversion (MEDI) algorithm6'8which generates a local susceptibility distribution by inversely modeling the estimated tissue field map, and incorporating prior information extracted from the magnitude images. The tissue field map was derived by removing the background field caused by significant susceptibility sources from the field map.

[0232] The DCEQP protocol involved seven pre-contrast T1 -weighted sequences with varying time delay (TD) values ranging from 120ms to 10 seconds. Each scan comprised the acquisition of five axial slices, positioned orthogonally to the internal carotid artery, facilitating later selection of an arterial input function. The dynamic scan consisted of 250 frames with a time resolution of 1.2 seconds.3,9MultiHance (gadobenate dimeglumine, Bracco Diagnostics, Inc.) was used at an appropriate dosage, the patient's weight was considered, half of that calculated volume was injected at a rate of 4 ml / s, precisely at 10.8 seconds into the dynamic scan.

[0233] Five axial T2-weighted turbo spin echo (TSE) images with high spatial resolution were obtained to serve as an anatomical reference for selecting lesion regions of interest (ROIs) using Image J software (Bethesda, MD, USA: US National Institutes of Health). DCEQP sequences were processed in MATLAB using a Patlak mathematical model to derive the permeability, cerebral blood flow, and the volume (i.e., CBF and CBV) maps.

[0234] Mean lesional permeability values were generated by selecting an region of interest around the lesion on the axial T2-weighted TSE slice that contained the largest lesion module including surrounding hypo-intense hemosiderin ring, then averaging the values of all pixels within the region from the CBKi map.9,10300224353.1 - 57 -

[0235] Plasma extraction

[0236] Blood samples were collected using standard clinical 10 mL heparinized vacutainer tubes. The plasma was isolated by centrifugation within two hours of collection, 8-10 mL of heparinized blood at 2000 X g at 4°C for 10 minutes. The supernatant plasma is divided in 12-14 aliquots of 300pl in 1.7ml microcentrifuge tubes and stored immediately at -80 C and processed in batches for proteins and metabolites assays.11,12

[0237] Proteins assays and quantification

[0238] Within each plate, the thawed aliquots of subjects’ plasma samples are placed in duplicate wells. To ensure external quality control, three quantikine immunoassay controls in diluted porcine serum were set up for each enzyme-linked immunosorbent assay biomarker (R& D systems, Minneapolis, Minnesota, USA).13These controls were loaded into duplicate wells. The control values from each plate were compared to the concentrations provided by the manufacturer. If a control value exceeds a variation precision established by the manufacturer, the plate will be rejected, and the experiment will be repeated.

[0239] A fully modular automated microplate wash and reagent dispenser system was used for accuracy and repeatability. A microplate reader set to 450-nm with a 570-nm wavelength correction BioRad BioPlex-100 analyzer (BioRad Laboratories, Hercules, California, USA) was used for measurements. These experiments follow the outlined guidelines by MarkVCID Consortium for consistency and reliability. A 4-parameter logistic regression analysis was employed to calculate the sample concentration. There was no batch effect using principal component analysis per day of assay.

[0240] Differentially expressed metabolites in discovery diagnostic and prognostic SH cohorts

[0241] Two propensity-matched cohorts of cerebral cavernous malformation (CCM) patients with and without SH were enrolled as part of separate biomarker discovery project (clinicaltrials.gov NCT04467489), independent of the 46 subjects in the current study. A diagnostic cohort of was constituted of 20 CCM patients who experienced SH during the year prior to blood collection and to 20 propensity-matched CCM patients without clinical events during the prior year. The patients were propensity-matched for (1) age at enrollment, (2) gender, (3) familial / sporadic-CM, (4) CM location (non-brainstem / brainstem), and (5) epilepsy (>1 seizure in the prior year, with / without medications).

[0242] The prognostic cohort included 15 CCM patients that developed CCM bleeding (SH or AC) during the year following blood collection, propensity-matched to 15 CCM patients300224353.1 - 58 -without bleeding during the following year. The patients were propensity-matched for (1) age at enrollment, (2) gender, (3) familial / sporadic-CCM, (4) CCM location (non-brainstem / brainstem), (5) epilepsy (>1 seizure in the prior year, with / without medications), and (6) prior SH in the year before enrolment. The diagnosis of CCM and associated clinical events were confirmed by an experienced senior neurosurgeon (IAA) through clinical MRI scans.

[0243] Sample extraction and processing for liquid chromatography-mass spectrometry assays

[0244] Fifty-pl of plasma were processed and analyzed using ultra-high performance liquid chromatography-tandem mass spectrometry (UPLC-MS / MS). Each sample was first chemically separated by polarity on a UPLC gradient and then analyzed by the mass detector in full scan mode (MS1). A pooled sample was used to collect iterative tandem MS (MS2) data. MS1data were used to create feature lists and assess relative abundances of metabolites while MS2data were used to identify the plasma compounds using fragmentation patterns. Once features were selected from the full feature list using statistical analysis, identification of selected features was performed through a combination of assigning molecular formula using the exact mass, molecular networking of fragmentation patterns using the GNPS platform (https: / / gnps.ucsd.edu / ProteoSAFe / static / gnps-splash.jsp), and in silico predictions of molecular identity using Sirius (https: / / bio.informatik.uni-jena.de / software / sirius).

[0245] Metabolites were extracted using extraction solvent (100% methanol spiked with heavy labeled internal standards and stored at -80°C) added to each plasma sample at a ratio of 50pL of plasma per 400pL of extraction solvent in microcentrifuge tubes. Samples were then vortexed for 30 seconds and centrifuged at -10°C, 20,000xg for 15min and the supernatant was subsequently used for metabolomic analysis. Seventy-five-pL of metabolite extract was added to prelabeled microcentrifuge tubes and dried down completely under vacuum using a Genevac EZ-2 Elite (Genevac Ltd., Ipswich, UK). Samples were resuspended in 200 pL of 5% acetonitrile, and then added to a thermomixer (Eppendorf, Hamburg, Germany) to resuspend analytes at 4 °C, 1000 rpm for 15 min with an infinite hold at 4 °C. Samples were finally centrifuged at 4°C, 20,000 x g for 15 min to remove insoluble debris and lOOpL of supernatant was transferred to a prelabeled mass spectrometry (MS) vial.

[0246] Samples were analyzed on Thermo Fisher ultra-performance liquid chromatography system coupled to an Orbitrap IQ-X mass spectrometer, operating in positive mode. Five-pL sample was injected onto a Cortecs© UPLC T3 Column (1.2 pm, 2.1 x 100 mm) fitted with Cortecs© UPLC T3 guard at 30 °C. The mobile phase A was water with 5% acetonitrile and 0.1% formic acid and mobile phase B was 100% acetonitrile with 0.1% formic acid. Gradient300224353.1 - 59 -elution started with 0% B with a flow rate of 0.40 mL / min for 0.2 min and linearly increased to 100% B over 10 min and these conditions were held constant for 1.0 min. Finally, reequilibration at 0% B was performed for 1.5 min. The electrospray ionization conditions were set with the spray voltage at 3.4 kV, vaporizer temp at 400°C, and detection window set to 120-1500 m / z. Internal mass calibration was performed with continuous infusion of Easy-IC© at the MS1level. A Deep Scan AcquireX data acquisition workflow was performed in which precursor selection for MS2scans was set to 120-1500 m / z with targeted mass list created by repeated injection of a pooled QC sample and a targeted exclusion list created by injection of a method blank. The isolation window was 0.7 m / z with no offset with stepped collision energies of 20, 40, 60, and 80 V combined into one MS2spectrum.

[0247] Raw data were processed in Mzmine 2.53 to produce a list of features (an m / z at a particular retention time) with the normalized peak area of that feature found in each sample. Noise was filtered out for mass detection by setting intensity cutoffs. Only signals above 4E4 in the MS1level and 1E4 in the MS2level were included, m / z and retention time tolerances along with peak shape criteria were set to extract individual peaks. The IQX has high mass and retention time accuracy, so tolerances are relatively strict (0.005 m / z and 0.10 min). Features within m / z and retention time tolerances that likely represent isotopologues were grouped together rather than considered two different features. Raw peak areas were normalized using a weighted contribution of all selected internal standards, weighted by distance. The distance of the standard peak to the peak being normalized was calculated as distance = (MZ difference) + (RT difference). All features with associated MS2scans were used to create a subset of features to be analyzed in GNPS. All MS2scans for a single feature were averaged to create a consensus spectrum for each feature. The cut-off for removing features based on Polled QC sample is to remove any feature with a % coefficient of variation (%CV) greater than 20%.

[0248] Subclass networking analyses

[0249] GNPS is used to analyze the collected MS2 data. GNPS (https: / / gnps.ucsd.edu) is an open-access platform that contains libraries of tandem mass spectrometry data.14Data analyzed in GNPS is not public until uploaded by the user to MassIVE for community use. The platform allows users to search their data for fragmentation patterns that match or are analogous to known library fragmentation patterns, and simultaneously create networks of related fragmentation patterns in their dataset. Different workflows are available in GNPS: Feature Networking uses data that has been curated into feature lists and the classical Molecular Networking uses unprocessed, raw data files. However, the latter workflow can include data that gets filtered out during feature list curation and MS2 spectra are combined into consensus300224353.1 - 60 -spectra in a different manner than during feature list curation with can result in idiosyncrasies between the two methods.

[0250] MS data were first processed with MZMINE215and the results were exported to GNPS for FBMN analysis. A molecular network was created with the Feature-Based Molecular Networking (FBMN) workflow16on GNPS (https: / / gnps.ucsd.edu).17Data were filtered by removing all MS / MS fragment ions within + / - 17 Da of the precursor m / z. MS / MS spectra were window filtered by choosing only the top six fragment ions in the + / - 50 Da window throughout the spectrum. The precursor ion mass tolerance was set to 0.02 Da and the MS / MS fragment ion tolerance to 0.02 Da. A molecular network was then created where edges were filtered to have a cosine score above 0.7 and more than four matched peaks. Further, edges between two nodes were kept in the network if and only if each of the nodes appeared in each other’s respective top 10 most similar nodes.

[0251] Finally, the maximum size of a molecular family was set to 100, and the lowest scoring edges were removed from molecular families until the molecular family size was below this threshold. The analogue search mode was used by searching against MS / MS spectra with a maximum difference of 100.0 in the precursor ion value. The library spectra were filtered in the same manner as the input data. All matches kept between network spectra and library spectra were required to have a score above 0.7 and at least four matched peaks. The DEREPLICATOR18was used to annotate MS / MS spectra. The molecular networks were visualized using Cytoscape software (http: / / www.cytoscape.org / ).19

[0252] Twelve non-correlated metabolites (Supplementary Table 2) that constituted the best biomarker models in both prognostic and diagnostic cohorts were selected to be assessed in the present study cohort. The best models selected had the lowest Akaike information criterion using in logistic regression, the higher penalty with Lasso penalty (LI) and coefficients having z-score >2 (p<0.05) in conditional logistic regression coefficient with standardized data.

[0253] Multi-omic integrative analyses of metabolome, proteome, microbiome and transcriptome

[0254] A multi-omic integrative analyses of the differential plasma metabolome, proteome, microbiome and transcriptome of CASH patients was further performed.20These additional analyses may illustrate the biological mechanisms reflected in the blood and detected by QSM and / or DCEQP.

[0255] The twelve metabolites differently expressed in the plasma of SH patients were queried in Comparative Toxicogenomics Database (CTD, http: / / ctdbase.org / ) in order to300224353.1 - 61 -identify interacting genes, enriched Kyoto Encyclopedia of Genes and Genomes (KEGG) and Reactome pathways (p<0.05, FDR corrected). In addition, the enriched-KEGG pathways (p<0.05, FDR corrected; Bayes Factor>3) associated with the differently expressed genes [p<0.05, FDR corrected] within the human lesional transcriptome of neurovascular units (NVUs) from CASH lesions were queried independently using Lynx (http: / / lynx.ci.uchicago.edu).21The common enriched-KEGG pathways between the differential metabolome and the lesional transcriptome were finally identified (p<0.05, FDR corrected; Bayes>3) using KEGG mapper (https: / / www.genome.jp / kegg / tool / map_pathwayl.html).

[0256] For the 16 candidate proteins, the coding genes of the plasma proteins were first queried using Uniprot (https: / / www.uniprot.org / ) and Genecards (https: / / www.genecards.org / ). KEGG, Reactome, BioCarta, WikiPathways, and BBID pathway analyses were then performed using DAVID (https: / / david.ncifcrf.gov / home.jsp).22The common enriched-KEGG and Reactome pathways between the differential plasma proteome and metabolome were identified. The genome of the bacterial gut species showing different relative abundances in CASH patients were queried using KEGG mapper (p<0.05, FDR corrected; Bayes>3).23The genes identified were further mapped into KEGG ortholog genes using AMON software24to extract reactions involving the ortholog genes and metabolites of interest.

[0257] Bias

[0258] The inventors discussed limitations of single site recruitment and referral biases in the discussion, but the inventors did not identify other potential confounders or potential biases in the analyses, since they applied to the same cohort. This is a sub-study of a larger Trials Readiness project1. Sample size was derived from the larger Trial Readiness study, and included the cohort of patients at the Chicago site where plasma samples were collected along with the imaging data.

[0259] Sensitivity analysis

[0260] Sensitivity analyses included co-correlations among individual biomarkers at individual points in time.

[0261] Supplementary Results

[0262] Correlations between blood biomarkers and imaging biomarkers at single timepoints

[0263] The preliminary univariate correlation analyses of the circulating biomarkers (i.e., proteins and metabolites) showed that only the plasma levels of angiopioetin-1 and300224353.1 - 62 -hypoxanthine were correlated ( =0.58). No significant correlation was observed between the plasma levels of any of the protein or metabolite assayed and the mean lesional QSM and / or DCEQP values at single points in time.

[0264] Change in proteins plasma levels vs. continuous change in mean QSM

[0265] The best combination of relative percent change in plasma levels proteins reflects the percent change in mean lesional QSM with 60.0% sensitivity and 94.0% specificity (SSE=3.06; =0.64 and / ?2=0.41; P= 3xl0’3) (FIG. 9A):QSM%~ 0.32*[LBP%] +0.30*[endogHn%] +0.1*[VEGF%] +0.04*[TM%] +9.8X10-3*[TLR4%] +5.6x10-4*[IL16%o] +2.3x10-5*[TSP1%] -O.15*[IL1B%] -0.15

[0266] The best combination of absolute change in proteins concentrations reflects with 33.0% sensitivity and 100% specificity the absolute change in mean lesional QSM (SSE=3.88; =0.26 and / ?2=0.07; P= 4.7xl0’2) (FIG. 9B):ΔQSM~ -2.9x10-4[ΔTSP1] -8.7x10-5*[AROB04]-2.0xlO-4*[ATLR4]-5.4*[AIL1B] -0.07

[0267] Change in proteins plasma levels vs. categorical change in mean QSM

[0268] A combination of absolute change in proteins plasma levels reflects the categorical change in mean lesional QSM>6% with 86.0% / 85.0% sensitivity / specificity [area under the curve (AUC) =90.2% (78.8%; 100%); P= 1.2X10’8] (FIG. 10):biQSM- 77.32*[AIL10] - 33.23*[AIL17] -0.68

[0269] Change in metabolites plasma levels vs. continuous change in mean QSM

[0270] Absolute change in acetyl. L. carnitine did correlate individually with absolute change in mean QSM and was the model of absolute change in proteins levels reflecting the absolute change in mean QSM with the highest documented correlation of =0.43 and R2=Q.18 (P=0.001) with sensitivity / specificity of 32.0% / 81.0% and a sum of squared error (SSP7) under LOOCV of 4.23:AQSM~ 0.07 *[Aacetyl. L.carnitineJ-0.03

[0271] The best combination of relative change in metabolites levels reflects the relative change in mean lesional QSM with 15.0% / 79.0% sensitivity / specificity (SSE=8.92; =0.20 and 7?2=0.04; P= 5.3xl0-1):QSM% ~ -0.10*[acetyl. L.carnitine%] -0.01*phenylacetylglutamine%] - 0.02*[piperine%]+ 0.01

[0272] Change in metabolites plasma levels vs. categorical change in mean QSM300224353.1 - 63 -

[0273] The best combination of absolute change in metabolites levels reflects with 75.0% sensitivity and 72.0% specificity the categorical change in mean lesional QSM>6% (AUC=80.5% [69.5%; 91.5%]; P= 0.004):biQSM ~ 0.08*[Aacetyl. L. carnitine] -0.32*[Apipecolic acid] -0.68

[0274] Change in proteins plasma levels vs. continuous change in mean DCEQP

[0275] The best combination of relative percent change in plasma levels proteins reflects the relative change in mean lesional DCEQP with 100% sensitivity and 24.0% specificity (SSE=47.61; p=0.25 and A2=0.04; P= 3.9xl0-1) including only two proteins:DCEQP% ~ -O.3O*[IL1B%] +0.07*[TLR4%] +0.82

[0276] The best combination of absolute change in proteins concentrations reflects with 96.0% sensitivity and 39.0% specificity the absolute change in mean lesional DCEQP (SSE=5.31; =0.06 and A2=0.11; P= 2.2X10’1):ADCEQP- -6.34*[AIL1B] -3.92x10-5*[A TSP1] +0.01*[A TSP2]-1.0x10-4*[AROBO4] +0.07

[0277] Change in proteins plasma levels vs. categorical change in mean DCEQP

[0278] The combination of relative percent change in proteins levels reflecting had 82.0% sensitivity and 81.0% specificity to detect the categorical change in mean lesional DCEQP>40% with highest / ! f / C =92.2% [85.9%; 98.5%]; P= 2.7X1O’10):biDCEQP~ 2.0*[endoglin%] -1.73*[VEGF%] +0.6*[IL16%] -0.86

[0279] Change in metabolites plasma levels vs. continuous change in mean DCEQP

[0280] The best combination of relative percent change in plasma levels metabolites reflects the relative change in mean lesional DCEQP with 100% sensitivity and 8.0% specificity (SSE=45.48; p=0.15 and A2=0.01; P= 8.6X10’1):DCEQP% ~ -0.06*[hypoxanthine%] -0.09*[piperine%]+0.90

[0281] The best combination of absolute change in metabolites levels reflects with 93.0% sensitivity and 21.0% specificity the absolute change in mean lesional DCEQP (SSE=6.24; p= -0.20 and A2=0.04; P= 3.1X10’1):ADCEQP~ 0.12*[ALPE 18:0] +0.01*[Apiperine ]+0.05

[0282] Change in metabolites plasma levels vs. categorical change in mean DCEQP300224353.1 - 64 -

[0283] The weighted combination of absolute change in plasma levels of metabolites reflect with 96.0% sensitivity and 91.0% specificity a categorical change in mean lesional DCEQP>40% (AUC=98.0% [95.4%; 100%]; P= 1.4xl0’15):biDCEQP ~ 0.30*[ALPE 18.0] -0.17*[Apiperine] - 0.04*[Apipecolic acid] -0.73

[0284] Change in proteins and metabolites plasma levels combinations vs. continuous change in mean DCEQP.

[0285] The composite model of relative percent change in proteins and metabolites plasma levels reflects the relative change in mean lesional DCEQP with 100% sensitivity and 25.0% specificity (SSE=53.07; p=0.39 and R2= -0.01; P= 4.6xl0-1):DCEQP% ~ -0.27*[ROB04%] - O.34*[IL1B%] -O.48*[acetyLL.carnitine%o]+1.47

[0286] The best combination of absolute change in proteins and metabolites levels reflects with 96.0% sensitivity and 46.0% specificity the absolute change in mean lesional DCEQP (SSE=5.25; =0.08 and A2=0.13; P= 2.8X10’1)):ADCEQP- 0.01*ATSP2] -l.lxl0-4*[A ROBO4] -1.43*[AILlB]-0.10*[Ahypoxanthine] -4.07x10-8*[ΔTSP1] +0.08* * *

[0287] All of the methods disclosed and claimed herein can be made and executed without undue experimentation in light of the present disclosure. While the compositions and methods of this invention have been described in terms of preferred embodiments, it will be apparent to those of skill in the art that variations may be applied to the methods and in the steps or in the sequence of steps of the method described herein without departing from the concept, spirit and scope of the invention. More specifically, it will be apparent that certain agents which are both chemically and physiologically related may be substituted for the agents described herein while the same or similar results would be achieved. All such similar substitutes and modifications apparent to those skilled in the art are deemed to be within the spirit, scope and concept of the invention as defined by the appended claims.REFERENCES

[0288] The following references and the references cited throughout the specification, to the extent that they provide exemplary procedural or other details supplementary to those set forth herein, are specifically incorporated herein by reference.300224353.1 - 65 -

[0289] Hage, S. et al. J Cereb Blood Flow Metab. 2025 Jun;45(6): 1153-1165. doi: 10.1177 / 0271678X251314366. Epub 2025 Jan 20.References for Example 21. Wang M, Carver JJ, Phelan VV, Sanchez LM, Garg N, Peng Y, Nguyen DD, Watrous J, Kapono CA, Luzzatto-Knaan T, Porto C, Bouslimani A, Melnik AV, Meehan MJ, Liu WT, et al. Sharing and community curation of mass spectrometry data with global natural products social molecular networking. VatBzotecAwo / .34:828-837; 2016.2. Pluskal T, Castillo S, Villar-Briones A, Oresic M. Mzmine 2: Modular framework for processing, visualizing, and analyzing mass spectrometry-based molecular profile data. BMC Bioinformatics.11:395; 2010.3. Nothias L-F, Petras D, Schmid R, Diihrkop K, Rainer J, Sarvepalli A, Protsyuk I, Ernst M, Tsugawa H, Fleischauer M. Feature-based molecular networking in the gnps analysis environment. Nature me / AotA.17:905-908; 2020.4. Wang M, Carver JJ, Phelan VV, Sanchez LM, Garg N, Peng Y, Nguyen DD, Watrous J, Kapono CA, Luzzatto-Knaan T. Sharing and community curation of mass spectrometry data with global natural products social molecular networking. Nature Z>zotecA«o / ogy.34:828-837; 2016.5. Mohimani H, Gurevich A, Shlemov A, Mikheenko A, Korobeynikov A, Cao L, Shcherbin E, Nothias L-F, Dorrestein PC, Pevzner PA. Dereplication of microbial metabolites through database search of mass spectra. Nature communications.9:4035; 2018.6. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T. Cytoscape: A software environment for integrated models of biomolecular interaction networks. Genome research.13:2498-2504; 2003.7. Polster SP, Cao Y, Carroll T, Flemming K, Girard R, Hanley D, Hobson N, Kim H, Koenig J, Koskimaki J, Lane K, Majersik JJ, McBee N, Morrison L, Shenkar R, et al. Trial readiness in cavernous angiomas with symptomatic hemorrhage (cash). Neurosurgery A 95 -964; 2019.8. Al-Shahi Salman R, Berg MJ, Morrison L, Awad IA. Hemorrhage from cavernous malformations of the brain: Definition and reporting standards. Angioma alliance scientific advisory board. Stroke.39:3222-3230,' 2008.9. Girard R, Fam MD, Zeineddine HA, Tan H, Mikati AG, Shi C, Jesselson M, Shenkar R, Wu M, Cao Y, Hobson N, Larsson HBW, Christoforidis GA, Awad IA. Vascular permeability and iron deposition biomarkers in longitudinal follow-up of cerebral cavernous malformations. J Neurosurg.127:102-110; 2017.300224353.1 - 66 -10. Tan H, Zhang L, Mikati AG, Girard R, Khanna O, Fam MD, Liu T, Wang Y, Edelman RR, Christoforidis G, Awad IA. Quantitative susceptibility mapping in cerebral cavernous malformations: Clinical correlations. AJNR Am J Neuroradiol3TA2Q9-Y2, 2016.11. Larsson HB, Courivaud F, Rostrup E, Hansen AE. Measurement of brain perfusion, blood volume, and blood-brain barrier permeability, using dynamic contrast-enhanced t(l)-weighted mri at 3 tesla. Magn Reson Med A2. \219-\2'A\ 2009.12. R Core Team R. R: A language and environment for statistical computing. 2013.References for Example 41. Akers A, Al-Shahi Salman R, Awad IA, et al. Synopsis of Guidelines for the Clinical Management of Cerebral Cavernous Malformations: Consensus Recommendations Based on Systematic Literature Review by the Angioma Alliance Scientific Advisory Board Clinical Experts Panel. Neurosurgery 2017;80(5):665-680; doi:10.1093 / neuros / nyx0912. Al-Shahi Salman R, Berg MJ, Morrison L, Awad IA, Angioma Alliance Scientific Advisory B. Hemorrhage from cavernous malformations of the brain: definition and reporting standards. Angioma Alliance Scientific Advisory Board. Stroke 2008;39(12):3222-3230; doi:10.1161 / STROKEAHA.108.5155443. Girard R, Khanna O, Shenkar R, et al. Peripheral plasma vitamin D and non-HDL cholesterol reflect the severity of cerebral cavernous malformation disease. Biomark Med 2016;10(3):255-264; doi:10.2217 / bmm.l5.1184. Carrion-Penagos J, Zeineddine HA, Polster SP, et al. Subclinical imaging changes in cerebral cavernous angiomas during prospective surveillance. J Neurosurg 2020; 134(3): 1147-1154; doi: 10.3171 / 2020. LJnsl934795. Girard R, Li Y, Stadnik A, et al. A Roadmap for Developing Plasma Diagnostic and Prognostic Biomarkers of Cerebral Cavernous Angioma With Symptomatic Hemorrhage (CASH). Neurosurgery 2021;88(3):686-697; doi:10.1093 / neuros / nyaa4786. Snellings DA, Hong CC, Ren AA, et al. Cerebral Cavernous Malformation: From Mechanism to Therapy. Circ Res 2021; 129(1): 195-215; doi:10.1161 / CIRCRESAHA.12L 3181747. Mikati AG, Khanna O, Zhang L, et al. Vascular permeability in cerebral cavernous malformations. J Cereb Blood Flow Metab 2015;35(10): 1632-1639; doi:10.1038 / jcbfm.2015.988. Tan H, Liu T, Wu Y, et al. Evaluation of iron content in human cerebral cavernous malformation using quantitative susceptibility mapping. Invest Radiol 2014;49(7):498-504; doi: 10.1097 / RLI.0000000000000043300224353.1 - 67 -9. Hage S, Kinkade S, Girard R, et al. Trial Readiness of Cavernous Malformations With Symptomatic Hemorrhage, Part II: Biomarkers and Trial Modeling. Stroke 2024;55(l):31-39; doi: 10.1161 / STROKEAHA.123.04408310. Girard R, Fam MD, Zeineddine HA, et al. Vascular permeability and iron deposition biomarkers in longitudinal follow-up of cerebral cavernous malformations. J Neurosurg 2017; 127(1): 102-110; doi:10.3171 / 2016.5. JNS1668711. Polster SP, Stadnik A, Akers AL, et al. Atorvastatin Treatment of Cavernous Angiomas with Symptomatic Hemorrhage Exploratory Proof of Concept (AT CASH EPOC) Trial. Neurosurgery 2019;85(6):843-853; doi:10.1093 / neuros / nyy53912. Srinath A, Xie B, Li Y, et al. Plasma metabolites with mechanistic and clinical links to the neurovascular disease cavernous angioma. Communications medicine 2023;3(l):35;13. Polster SP, Sharma A, Tanes C, et al. Permissive microbiome characterizes human subjects with a neurovascular disease cavernous angioma. Nat Commun 2020; 11(1):2659; doi: 10.1038 / s41467-020- 16436-w14. Girard R, Zeineddine HA, Koskimaki J, et al. Plasma Biomarkers of Inflammation and Angiogenesis Predict Cerebral Cavernous Malformation Symptomatic Hemorrhage or Lesional Growth. Circ Res 2018;122(12): 1716-1721; doi:10.1161 / CIRCRESAHA.l 18.31268015. Lyne SB, Girard R, Koskimaki J, et al. Biomarkers of cavernous angioma with symptomatic hemorrhage. JCI Insight 2019;4(12); doi: 10.1172 / j ci. insight.12857716. Sharygin D, Koniaris LG, Wells C, Zimmers TA, Hamidi T. Role of CD14 in human disease. Immunology 2023;169(3):260-270; doi: 10.1111 / imm.1363417. Long H, Zhu W, Wei L, Zhao J. Iron homeostasis imbalance and ferroptosis in brain diseases. MedComm (2020) 2023;4(4):e298; doi:10.1002 / mco2.29818. Lin S, Yin Q, Zhong Q, et al. Heme activates TLR4-mediated inflammatory injury via MyD88 / TRIF signaling pathway in intracerebral hemorrhage. J Neuroinflammation 2012;946; doi: 10.1186 / 1742-2094-9-4619. Xu J, Xiao C, Song W, et al. Elevated Heme Oxygenase-1 Correlates With Increased Brain Iron Deposition Measured by Quantitative Susceptibility Mapping and Decreased Hemoglobin in Patients With Parkinson's Disease. Front Aging Neurosci 2021;13656626; doi: 10.3389 / fnagi.202L 65662620. Rashdan NA, Zhai B, Lovern PC. Fluid shear stress regulates placental growth factor expression via heme oxygenase 1 and iron. Sci Rep 2021; 11(1): 14912; doi:10.1038 / s41598-021-94559-w300224353.1 - 68 -21. Shi C, Shenkar R, Du H, et al. Immune response in human cerebral cavernous malformations. Stroke 2009;40(5): 1659-1665; doi: 10.1161 / STROKEAHA.108.538769 22. Ramakrishnan S, Anand V, Roy S. Vascular endothelial growth factor signaling in hypoxia and inflammation. J Neuroimmune Pharmacol 2014;9(2): 142-160; doi:10.1007 / sl 1481-014-9531-723. Huang L, Fu L. Mechanisms of resistance to EGFR tyrosine kinase inhibitors. Acta Pharm Sin B 2015' 5(5.390-401' doi:10.1016 / j.apsb.2015.07.00124. Shi F, Li Q, Liu S, et al. Porcine circovirus type 2 upregulates endothelial-derived IL-8 production in porcine iliac artery endothelial cells via the RIG-I / MDA-5 / MAVS / JNK signaling pathway. BMC Vet Res 2020;16(l):265; doi:10.1186 / sl2917-020-02486-l25. Wu Z, Zhang Z, Lei Z, Lei P. CD14: Biology and role in the pathogenesis of disease. Cytokine & growth factor reviews 2019;4824-31;26. Conway EM. Thrombomodulin and its role in inflammation. Seminars in immunopathology; 2012: Springer: 107-125.27. Jones CA, London NR, Chen H, et al. Robo4 stabilizes the vascular network by inhibiting pathologic angiogenesis and endothelial hyperpermeability. Nature medicine 2008;14(4):448-453;28. Kumar V. Macrophages: the potent immunoregulatory innate immune cells.. In: Bhat KH, ed. Macrophage Activation-Biology and Disease'. Intechopen, 2019: 1-30.29. Lendeckel U, Venz S, Wolke C. Macrophages: shapes and functions. ChemTexts 2022;8(2): 12; doi: 10.1007 / s40828-022-00163-430. Burks S, Raymick J, Robinson B, Hanig J, Sarkar S. Neuroprotective effects of acetyl-1-carnitine (ALC) in a chronic MPTP-induced Parkinson's disease mouse model: Endothelial and microglial effects. Neurosci Ze 2019;70386-95; doi:10.1016 / j.neulet.2019.03.015 31. Lal A, Atamna W, Killilea DW, Suh JH, Ames BN. Lipoic acid and acetyl-camitine reverse iron-induced oxidative stress in human fibroblasts. Redox Rep 2008;13(l):2-10; doi:10.1179 / 135100008x25915032. Baci D, Bruno A, Bassani B, et al. Acetyl-l-carnitine is an anti-angiogenic agent targeting the VEGFR2 and CXCR4 pathways. Cancer Lett 2018;429100-116; doi:10.1016 / j.canlet.2018.04.01833. Elliott GR, Lauwen AP, Bonta IL. The effect of acute feeding of carnitine, acetyl carnitine and propionyl carnitine on basal and A23187-stimulated eicosanoid release from rat carrageenan-elicited peritoneal macrophages. Br J Nutr 1990;64(2):497-503; doi: 10.1079 / bjn 19900049300224353.1 - 69 -34. Cunha SI, Magnusson PU, Dejana E, Lampugnani MG. Deregulated TGF-p / BMP Signaling in Vascular Malformations. Circ Res 2017;121(8):981-999; doi: 10.1161 / circresaha.117.30993035. Chen Y, Hao Q, Kim H, et al. Soluble endoglin modulates aberrant cerebral vascular remodeling. Ann Neurol 2009;66(l): 19-27; doi:10.1002 / ana.2171036. Lebrin F, Goumans MJ, Jonker L, et al. Endoglin promotes endothelial cell proliferation and TGF-beta / ALKl signal transduction. EMBO J 2004;23(20):4018-4028; doi: 10.1038 / sj.emboj.760038637. Lee NY, Blobe GC. The interaction of endoglin with beta-arrestin2 regulates transforming growth factor-beta-mediated ERK activation and migration in endothelial cells. The Journal of biological chemistry 2007;282(29):2l 507-21517; doi: 10.1074 / jbc. M70017620038. Armstrong LC, Bjorkblom B, Hankenson KD, Siadak AW, Stiles CE, Bornstein P. Thrombospondin 2 inhibits microvascular endothelial cell proliferation by a caspaseindependent mechanism. MolBiol Cell 2002;13(6):1893-1905; doi:10.1091 / mbc.e01-09-0066 39. Lopez-Ramirez MA, Fonseca G, Zeineddine HA, et al. Thrombospondin 1 (TSP1) replacement prevents cerebral cavernous malformations. J Exp Med 2017;214(l 1 ): 3331-3346; doi: 10.1084 / jem.2017117840. Carbone ML, Failla CM. Interleukin role in the regulation of endothelial cell pathological activation. Vase Biol 2021;3(l): R96-rl05; doi:10.1530 / vb-21-001041. Kim YJ, Ryu HM, Choi JY, et al. Hypoxanthine causes endothelial dysfunction through oxidative stress-induced apoptosis. Biochem Biophys Res Commun 2017;482(4):821-827; doi: 10.1016 / j.bbrc.2016.11.11942. Lopez-Ramirez MA, Lai CC, Soliman SI, et al. Astrocytes propel neurovascular dysfunction during cerebral cavernous malformation lesion formation. J Clin Invest 2021; 131(13); doi: 10.1172 / j ci 13957043. Doucette CD, Hilchie AL, Liwski R, Hoskin DW. Piperine, a dietary phytochemical, inhibits angiogenesis. J Nutr Biochem 2013;24(l):231-239; doi:10.1016 / j.jnutbio.2012.05.009 44. Bae GS, Kim MS, Jung WS, et al. Inhibition of lipopolysaccharide-induced inflammatory responses by piperine. Eur J Pharmacol 2010;642(l-3):154-162; doi:10.1016 / j.ejphar.2010.05.02645. Moore KN, Sill MW, Tenney ME, et al. A phase II trial of trebananib (AMG 386; IND#111071), a selective angiopoietin 1 / 2 neutralizing peptibody, in patients with300224353.1 - 70 -persistent / recurrent carcinoma of the endometrium: An NRG / Gynecologic Oncology Group trial. Gynecol Oncol 2015; 138(3):513-518; doi: 10.1016 / j.ygyno.2015.07.00646. Zhao Y, Li D, Liu R, Yuan Y. Bayesian optimal phase II designs with dual-criterion decision making. Pharm Stat 2023;22(4):605-618; doi:10.1002 / pst.2296References for Example 51. Hage S, Kinkade S, Girard R, et al. Trial Readiness of Cavernous Malformations With Symptomatic Hemorrhage, Part IL Biomarkers and Trial Modeling. Stroke. Jan 2024; 55( 1 ): 31 -39. doi:10.1161 / STROKEAHA.123.0440832. Lyne SB, Girard R, Koskimaki J, et al. Biomarkers of cavernous angioma with symptomatic hemorrhage. JCI Insight. Jun 202019;4(12)doi: 10.1172 / j ci. insight.128577 3. Girard R, Fam MD, Zeineddine HA, et al. Vascular permeability and iron deposition biomarkers in longitudinal follow-up of cerebral cavernous malformations. J Neurosurg. Jul 2017; 127(1): 102-110. doi:10.3171 / 2016.5. JNS166874. Tan H, Zhang L, Mikati AG, et al. Quantitative Susceptibility Mapping in Cerebral Cavernous Malformations: Clinical Correlations. AJNR Am J Neuroradiol. Jul 2016; 37(7): 1209- 15. doi: 10.3174 / aj nr. A47245. Zeineddine HA, Girard R, Cao Y, et al. Quantitative susceptibility mapping as a monitoring biomarker in cerebral cavernous malformations with recent hemorrhage. J Magn Re son Imaging. Apr 2018;47(4): 1133-1138. doi: 10.1002 / jmri.258316. Liu T, Khalidov I, de Rochefort L, et al. A novel background field removal method for MRI using projection onto dipole fields (PDF). NMR Biomed. Nov 2011;24(9): 1129-36. doi:10.1002 / nbm,16707. Liu T, Liu J, de Rochefort L, et al. Morphology enabled dipole inversion (MEDI) from a single-angle acquisition: comparison with COSMOS in human brain imaging. Magn Reson Med. Sep 2011;66(3):777-83. doi:10.1002 / mrm.228168. Liu T, Wisnieff C, Lou M, Chen W, Spincemaille P, Wang Y. Nonlinear formulation of the magnetic field to source relationship for robust quantitative susceptibility mapping. Magn Reson Med. Feb 2013;69(2):467-76. doi: 10.1002 / mrm.242729. Mikati AG, Tan H, Shenkar R, et al. Dynamic permeability and quantitative susceptibility: related imaging biomarkers in cerebral cavernous malformations. Stroke. Feb 2014;45(2):598-601. doi: 10.1161 / STROKEAHA.113.00354810. Larsson HB, Courivaud F, Rostrup E, Hansen AE. Measurement of brain perfusion, blood volume, and blood-brain barrier permeability, using dynamic contrast-enhanced T(l)-weighted MRI at 3 tesla. Magn Reson Med. Nov 2009;62(5): 1270-81. doi: 10.1002 / mrm.22136300224353.1 - 71 -11. Girard R, Zeineddine HA, Fam MD, et al. Plasma Biomarkers of Inflammation Reflect Seizures and Hemorrhagic Activity of Cerebral Cavernous Malformations. Transl Stroke Res. Feb 2018;9(l):34-43. doi:10.1007 / sl2975-017-0561-312. Girard R, Zeineddine HA, Koskimaki J, et al. Plasma Biomarkers of Inflammation and Angiogenesis Predict Cerebral Cavernous Malformation Symptomatic Hemorrhage or Lesional Growth. Circ Res. Jun 8 2018; 122(12): 1716-1721. doi:10.1161 / CIRCRESAHA.l 18.31268013. Zhou X, Fragala MS, McElhaney JE, Kuchel GA. Conceptual and methodological issues relevant to cytokine and inflammatory marker measurements in clinical research. Curr Opin Clin Nutr Metab Care. Sep 2010;13(5):541-7. doi:10.1097 / MCO.0b013e32833cf3bc 14. Wang M, Carver JJ, Phelan VV, et al. Sharing and community curation of mass spectrometry data with Global Natural Products Social Molecular Networking. NatBiotechnol. Aug 92016;34(8):828-837. doi:10.1038 / nbt.359715. Pluskal T, Castillo S, Villar-Briones A, Oresic M. MZmine 2: modular framework for processing, visualizing, and analyzing mass spectrometry-based molecular profile data. BMC Bioinformatics. Jul 23 2010;l 1:395. doi: 10.1186 / 1471-2105-11-39516. Nothias L-F, Petras D, Schmid R, et al. Feature-based molecular networking in the GNPS analysis environment. Nature methods. 2020;17(9):905-908.17. Wang M, Carver JJ, Phelan VV, et al. Sharing and community curation of mass spectrometry data with Global Natural Products Social Molecular Networking. Nature biotechnology. 2016;34(8): 828-837.18. Mohimani H, Gurevich A, Shlemov A, et al. Dereplication of microbial metabolites through database search of mass spectra. Nature communications. 2018;9(l):4035.19. Shannon P, Markiel A, Ozier O, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome research. 2003;13(l l):2498-2504. 20. Koskimaki J, Polster SP, Li Y, et al. Common transcriptome, plasma molecules, and imaging signatures in the aging brain and a Mendelian neurovascular disease, cerebral cavernous malformation. Geroscience. Oct 2020;42(5):1351-1363. doi:10.1007 / sll357-020-00201-421. Sulakhe D, Balasubramanian S, Xie B, et al. Lynx: a database and knowledge extraction engine for integrative medicine. Nucleic Acids Res. Jan 2014;42(Database issue): D1007-12. doi:10.1093 / nar / gktll6622. Huang DW, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nature protocols. 2009;4(l):44-57.300224353.1 - 72 -23. Polster SP, Sharma A, Tanes C, et al. Permissive microbiome characterizes human subjects with a neurovascular disease cavernous angioma. Nat Commun. May 27 2020; 11 ( 1 ): 2659. doi: 10.1038 / s41467-020- 16436-w24. Shaffer M, Thurimella K, Quinn K, et al. AMON: annotation of metabolite origins via networks to integrate microbiome and metabolome data. BMC Bioinformatics. Nov 28 2019;20(l):614. doi:10.1186 / sl2859-019-3176-8300224353.1 - 73 -

Claims

WHAT IS CLAIMED:

1. A method of measuring biomarkers in a patient having, suspected of having, or diagnosed with having cavernous angioma (CA), the method comprising measuring an amount of one or more biomarkers in a biological sample from the patient, wherein the one or more biomarkers comprise tyrosine, acetyl-L-camitine, arachidonic acid, hexanoylcamitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, vascular endothelial growth factor-A (VEGF), interleukin (IL)- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, thrombospondin- 1 (TSP1), thrombospondin-2 (TSP2), thrombomodulin (TM), cluster of differentiation- 14 (CD 14), c-reactive protein, roundabout guidance receptor 4 (ROBO4), lipopolysaccharide binding protein (LBP), or a combination thereof.

2. The method of claim 1, wherein the one or more biomarkers comprise ROBO4, CD14, TM, and acetyl-L-camitine in the biological sample.

3. The method of claim 1, wherein the one or more biomarkers comprise ROBO4, CD14, and TM in the biological sample.

4. The method of claim 1, wherein the one or more biomarkers comprise endoglin, pipecolic acid, arachidonic acid, and hypoxanthine in the biological sample.

5. The method of claim 1, wherein the one or more biomarkers comprise endoglin, TSP2, and IL-16 in the biological sample.

6. The method of any one of claims 1 to 5, wherein one or more of the biomarkers are determined to have differential levels relative to a control.

7. The method of claim 6, wherein one or more of the biomarkers are determined to have levels significantly different than the control.

8. The method of claim 6, wherein one or more of the biomarkers are determined to have levels not significantly different than the control.

9. The method of any one of claims 6 to 8, wherein the control comprises the level of the biomarker(s) in samples from subjects identified as non-CASH, low risk, as not having CA, or as having non-aggressive CA.

10. The method of any one of claims 6 to 9, wherein the control comprises the level of the biomarker(s) in samples from subjects identified as CASH, high risk, as having CA, or as having aggressive CA.

11. The method of any one of claims 6 to 8, wherein the control comprises the level of the biomarker(s) in a second biological sample obtained from the patient prior to the measuring.300224353.1 - 74 -12. The method of any one of claims 6 to 11, wherein the measuring is performed on two or more biological samples obtained from the patient.

13. The method of claim 12, wherein the two or more biological samples are obtained at different times.

14. The method of any one of claims 1 to 13, wherein the CA is further defined as CA with symptomatic hemorrhage (CASH).

15. The method of any one of claims 1 to 14, wherein the CA is further defined as CA without symptomatic hemorrhage (non-CASH).

16. The method of any one of claims 1 to 15, wherein the sample from the subject comprises a tissue sample, a blood sample, a whole blood sample, a fractionated sample, a plasma sample, a fecal sample, or a urine sample.

17. The method of any one of claims 1 to 16, wherein the sample from the subject comprises a plasma sample from the subject.

18. The method of any of claims 1 to 17, wherein the expression level of no other biomarker in the sample is measured.

19. The method of any one of claims 1 to 18, further comprising administering a treatment to the patient.

20. The method of claim 19, wherein the treatment comprises a gene therapy, a B-cell immunomodulation therapy, surgical excision of a CCM lesion, or combinations thereof.

21. A method for treating cavernous angioma (CA) or CASH in a patient, the method comprising administering a therapy to a subject that has been evaluated for one or more biomarkers in a biological sample from the patient, wherein the one or more biomarkers comprise tyrosine, acetyl-L-camitine, arachidonic acid, hexanoylcarnitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, vascular endothelial growth factor- A (VEGF), interleukin (IL)- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, thrombospondin- 1 (TSP1), thrombospondin-2 (TSP2), thrombomodulin (TM), cluster of differentiation- 14 (CD14), c-reactive protein, roundabout guidance receptor 4 (ROBO4), and lipopolysaccharide binding protein (LBP), or a combination thereof.

22. The method of claim 21, wherein the one or more biomarkers comprise ROBO4, CD14, TM, and acetyl-L-camitine in the biological sample.

23. The method of claim 21, wherein the one or more biomarkers comprise ROBO4, CD14, and TM in the biological sample.300224353.1 - 75 -24. The method of claim 21, wherein the one or more biomarkers comprise endoglin, pipecolic acid, arachidonic acid, and hypoxanthine in the biological sample.

25. The method of claim 21, wherein the one or more biomarkers comprise endoglin, TSP2, and IL-16 in the biological sample.

26. The method of any one of claims 21 to 25, wherein one or more of the biomarkers are determined to have a differential level relative to a control.

27. The method of claim 26, wherein one or more of the biomarkers are determined to have levels significantly different than the control.

28. The method of claim 26, wherein one or more of the biomarkers are determined to have levels not significantly different than the control.

29. The method of any one of claims 26 to 28, wherein the control comprises the level of the biomarker in samples from subjects identified as non-CASH, low risk, as not having CA, or as having non-aggressive CA.

30. The method of any one of claims 26 to 29, wherein the control comprises the level of the biomarker in samples from subjects identified as CASH, high risk, as having CA, or as having aggressive CA.

31. The method of any one of claims 26 to 28, wherein the control comprises the level of the biomarker(s) in a second biological sample obtained from the patient prior to the measuring.

32. The method of any one of claims 26 to 31, wherein the measuring is performed on two or more biological samples obtained from the patient.

33. The method of claim 32, wherein the two or more biological samples are obtained at different times.

34. The method of any one of claims 21 to 33, wherein the CA is further defined as CA with symptomatic hemorrhage (CASH).

35. The method of any one of claims 21 to 34, wherein the CA is further defined as CA without symptomatic hemorrhage (non-CASH).

36. The method of any one of claims 21 to 35, wherein the sample from the subject comprises a tissue sample, a blood sample, a whole blood sample, a fractionated sample, a plasma sample, a fecal sample, or a urine sample.

37. The method of any one of claims 21 to 36, wherein the sample from the subject comprises a plasma sample from the subject.

38. The method of any of claims 21 to 37, wherein the expression level of no other biomarker in the sample was evaluated.300224353.1 - 76 -39. The method of any one of claims 21 to 38, wherein the treatment comprises a gene therapy, a B-cell immunomodulation therapy, surgical excision of a CCM lesion, or combinations thereof.

40. A method for prognosing, diagnosing, or monitoring CA, aggressive CA, non-aggressive CA, CASH, non-CASH, familial CA, or sporadic CA in a subject, the method comprising measuring the amount of one or more biomarkers in a biological sample from the patient, wherein the one or more biomarkers comprise tyrosine, acetyl-L-carnitine, arachidonic acid, hexanoylcamitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, vascular endothelial growth factor- A (VEGF), interleukin (IL)- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, thrombospondin- 1 (TSP1), thrombospondin-2 (TSP2), thrombomodulin (TM), cluster of differentiation- 14 (CD 14), c-reactive protein, roundabout guidance receptor 4 (ROBO4), and lipopolysaccharide binding protein (LBP), or a combination thereof.

41. The method of claim 40, wherein the one or more biomarkers comprise ROBO4, CD14, TM, and acetyl-L-camitine in the biological sample.

42. The method of claim 40, wherein the one or more biomarkers comprise ROBO4, CD14, and TM in the biological sample.

43. The method of claim 40, wherein the one or more biomarkers comprise endoglin, pipecolic acid, arachidonic acid, and hypoxanthine in the biological sample.

44. The method of claim 40, wherein the one or more biomarkers comprise endoglin, TSP2, and IL-16 in the biological sample.

45. The method of any one of claims 40 or 44, wherein one or more of the biomarkers are determined to have a differential level relative to a control.

46. The method of claim 45, wherein one or more of the biomarkers are determined to have levels significantly different than the control.

47. The method of claim 45, wherein one or more of the biomarkers are determined to have levels not significantly different than the control.

48. The method of any one of claims 45 to 47, wherein the control comprises the level of the biomarker in samples from subjects identified as non-CASH, low risk, as not having CA, or as having non-aggressive CA.

49. The method of any one of claims 45 to 48, wherein the control comprises the level of the biomarker in samples from subjects identified as CASH, high risk, as having CA, or as having aggressive CA.300224353.1 - 77 -50. The method of any one of claims 45 to 47, wherein the control comprises the level of the biomarker(s) in a second biological sample obtained from the patient.

51. The method of claim 50, wherein the second biological sample is obtained at a different time from obtaining the biological sample.

52. The method of any one of claims 40 to 51, wherein the CA is further defined as CA with symptomatic hemorrhage (CASH).

53. The method of any one of claims 40 to 52, wherein the CA is further defined as CA with symptomatic hemorrhage (CASH).

54. The method of any one of claims 40 to 52, wherein the CA is further defined as CA without symptomatic hemorrhage (non-CASH).

55. The method of any one of claims 40 to 54, wherein the sample from the subject comprises a tissue sample, a blood sample, a whole blood sample, a fractionated sample, a plasma sample, a fecal sample, or a urine sample.

56. The method of any one of claims 40 to 55, wherein the sample from the subject comprises a plasma sample from the subject.

57. The method of any one of claims 40 to 56, wherein the expression level of no other biomarker in the sample was measured.

58. The method of any one of claims 40 to 57, further comprising administering a treatment to the patient.

59. The method of claim 14, wherein the treatment comprises a gene therapy, a B-cell immunomodulation therapy, surgical excision of a CCM lesion, or combinations thereof.

60. A kit comprising detection agents for determining levels of one or more biomarkers, wherein the biomarkers comprise tyrosine, acetyl-L-carnitine, arachidonic acid, hexanoylcarnitine, hypoxanthine, LPE 18:0, LPE 18:1, methionine, phenylacetylglutamine, pipecolic acid, piperine, tryptophan, hydroxyphenyllactic acid, vascular endothelial growth factor-A (VEGF), interleukin (IL)- 10, IL-1β, IL- 16, endoglin, angiopoietin-1, angiopoietin-2, thrombospondin- 1 (TSP1), thrombospondin-2 (TSP2), thrombomodulin (TM), cluster of differentiation- 14 (CD 14), c-reactive protein, roundabout guidance receptor 4 (ROBO4), and lipopolysaccharide binding protein (LBP), or a combination thereof.

61. The method of claim 60, wherein the one or more biomarkers comprise ROBO4, CD14, TM, and acetyl-L-carnitine.

62. The method of claim 60, wherein the one or more biomarkers comprise ROBO4, CD14, and TM.300224353.1 - 78 -63. The method of claim 60, wherein the one or more biomarkers comprise endoglin, pipecolic acid, arachidonic acid and hypoxanthine.

64. The method of claim 60, wherein the one or more biomarkers comprise endoglin, TSP2, and IL-16.

65. The kit of any one of claims 60 to 64, wherein the kit further comprises one or more negative or positive control samples and / or control detection agents.

66. The kit of any one of claims 60 to 65, wherein the kit further comprises instructions for use.

67. The kit of any one of claims 60 to 66, wherein the kit comprises reagents for isolation and / or amplification of biomarkers and / or biomarkers from a biological sample.

68. The kit of claim 67, wherein the biological sample comprises a tissue sample, a blood sample, a whole blood sample, a fractionated sample, a plasma sample, a fecal sample, or a urine sample.

69. The kit of claim 67, wherein the biological sample comprises a plasma sample.

70. The kit of any one of claims 67 to 69, wherein the biological sample is from a subject.

71. The kit of any one of claims 60 to 70, wherein the subject has been diagnosed with CA 72. The kit of any one of claims 60 to 71, wherein the kit excludes reagents for detection of any other biomarkers.300224353.1 - 79 -