CROSS REFERENCE TO RELATED APPLICATION

[0001] This application claims priority to U.S. provisional application no. 62/534,072, filed July 18, 2017, the disclosure of which is incorporated herein by reference. FIELD

[0002] The present disclosure relates generally to biomarkers for intracranial aneurysm. The present disclosure is also related to the determination of aneurysm size and location by medical imaging following the assessment of aneurysm presence. The disclosure also relates to determining that the individual is at risk of intracranial aneurysm rupture, and/or for subarachnoid hemorrhage (SAH).

BACKGROUND

[0003] Intracranial aneurysms (IAs) are potentially deadly lesions in the

cerebrovasculature that, if ruptured, cause devastating subarachnoid hemorrhages. Of the roughly 30,000 people in the United States per year who experience aneurysmal

subarachnoid hemorrhage, approximately half die within one month, while half of the survivors live with permanent disability. It is estimated that 5% of Americans harbor unruptured IAs, but the exact prevalence is unknown. Since the majority of unruptured aneurysms are completely asymptomatic, most remain dormant and undetected. Aside from incidental discoveries on medical imaging performed for other purposes, IAs are often only discovered after rupture.

[0004] Clearly, early detection of IAs before they rupture is critical, as it would allow for vigilant monitoring by medical imaging and preventive treatment. Despite surgical risks, recent studies have demonstrated that treatment of unruptured IAs is able to drastically reduce the rate of rupture. For 50-year-old males, the probability of rupture during the patient's remaining lifetime is 22.8%, but is reduced to 1.6% after surgical clipping or 3.4% after endovascular coiling. Furthermore, preventative treatment of IAs has been found to be cost-effective, increasing the net quality-adjusted life-year. Screening tools that can identify patients with unruptured IAs would thus represent a major advance for patient care, as the detection and subsequent management of IAs could drastically reduce their catastrophic consequences and associated healthcare costs.

[0005] Unfortunately, screening for individuals with unruptured IAs is problematic.

Stratifying risk for IA in the general population by environmental and genetic risk factors (e.g. age, female sex, and hypertension) does not reliably identify patients with unruptured aneurysm. These risk factors are often shared with other cardiovascular diseases and do not independently correlate with the presence of IA.

[0006] Unruptured IAs have been incidentally detected by magnetic resonance imaging (MRI), computed tomography angiography (CTA) or digital subtraction angiography (DSA), mostly performed for other reasons. Yet, as stated by the American Stroke

Association, these imaging procedures are not suitable for IA screening, because the potential risks associated with them (e.g. some modalities are invasive and can cause complications) are not yet justified, especially considering their high cost. Even in high aneurysm risk populations (e.g. with family history of IA), it is debated whether patients should be screened by imaging, as its cost-effectiveness has not been clearly demonstrated. This raises a critical need for an alternative strategy for aneurysm detection that is minimally-invasive, affordable, and reliable. The present disclosure is pertinent to this need.

SUMMARY

[0007] The present disclosure provides methods for analysis of aneurisms, and in particular, IAs. In embodiments, the disclosure provides for determining the expression of any combination of biomarkers described herein. In embodiment, the disclosure provides a method for determining a presence (or absence) of an IA in a subject by analyzing a biological sample from the subject for expression of a combination of biomarkers that are described in Example 1, Table 2, and/or Example 2, Table 3, of this disclosure. In

embodiments, the disclosure provides for determining the presence of an IA in an individual by determining a difference in expression of a combination of biomarkers described herein, relative to a control. In embodiments, determining that biomarkers as described herein are the same as a normal control indicates the individual does not have an IA. In embodiments, the disclosure provides for determining the presence of an IA by determining that expression of at least one of the following biomarkers is increased relative to a control: PVRL2, CYP1B1, CD177, PDE9A, ARMC12, OLAH , TGS1, CD163, LOC100506229, OCLN, SEMA6B, ADTRP, VWA8, MTR R2L1, HOXB2, EPCAM, and IL18R1, and/or by determining that expression of at least one of the following biomarkers is decreased relative to a control:

IGSF23, PTGES, G0S2, FCRL5, Clorf226, UTS2, HBG2, CYP26B1, and C1QL1.

[0008] In embodiments, the disclosure comprises determining that expression of at least one of the following biomarkers is increased relative to the control: PVRL2, PDE9A, TGS1, LOC100506229, OCLN, SEMA6B, MTRNR2L1, HOXB2, EPCAM, or IL18R1; and/or that expression of at least one of the following biomarkers is decreased relative to the control: IGSF23, PTGES, UTS2, HBG2, CYP26B1, or C1QL1.

[0009] In embodiments, expression of at least one of C21orfl5, CYP1B 1, FLT3,

XKR3, SLC12A7, PAM, GPC4, FBN1, IL-8, GBP5, ETV7, MFSD9, SERPING1, TCL1A and CARD 17 is analyzed.

[0010] In embodiments, the disclosure provides for determining increased expression of C21orfl5, CYP1B1, FLT3, XKR3, SLC12A7, and PAM, relative to a control, and/or determining decreased expression of GPC4, FBN1, and IL-8 relative to the control, to indicate the presence of an IA.

[0011] In embodiments, the disclosure comprises determining increased expression

C21orfl5, CYP1B1, FLT3, XKR3, SLC12A7, PAM, and TCL1A relative to the control, and/or determining decreased expression of GPC4, FBN1, IL-8, GBP 5, ETV7, MFSD9, SERPING1 and CARD 17 relative to the control, to determine the presence of the IA.

[0012] In embodiments, the disclosure comprises determining the presence of the IA in the individual based on expression of the biomarkers, and performing a medical procedure on the individual. In certain approaches, the medical procedure comprises imaging the aneurysm. In certain approaches, the imaging comprises determining the size and location of the intracranial aneurysm. In embodiments, the medical procedure comprises determining that the intracranial aneurysm is a fusiform intracranial aneurysm, and optionally includes treating the fusiform intracranial aneurysm with a flow diverter. In embodiments, the medical procedure comprises determining that the intracranial aneurysm is a saccular intracranial aneurysm, and optionally includes treating the saccular intracranial aneurysm by

endovascular coiling or surgical clipping.

[0013] In embodiments, the disclosure comprises repeating a method described herein to monitor the subject with respect to the size and/or location and/or a response to treatment of the intracranial aneurism.

[0014] In certain approaches, the disclosure comprises determining the presence of the IA in the individual, and further determining that the individual is at risk of having the IA rupture, and/or for subarachnoid hemorrhage (SAH). BRIEF DESCRIPTION OF FIGURES

[0015] Figure 1. Expression differences between patients with IAs and controls, and an IA-associated expression signature. The volcano plot demonstrates differential RNA expression between the two groups. Grey circles indicate an IA-associated signature of significantly differentially expressed transcripts (p<0.05) with an absolute fold-change>2.

[0016] Figure 2. Verification of RNA expression differences by RT-qPCR.

Quantitative PCR performed on 5 prominent differentially expressed transcripts demonstrates that both the magnitude and direction of the fold-change in expression measured by RNA sequencing are similar to that measured by qPCR. Only 21 samples were analyzed by RT- qPCR because one control sample did not have enough RNA for the additional reactions. (Negative fold-change values calculated by negative inverse of fold-change, data

points=average values, error bars=standard error.)

[0017] Figure 3. Dimensionality reduction analyses separate blood samples from

IA patients and controls. (A) Principal component analysis (PCA) using all transcriptome data demonstrates aggregation of samples from IA patients (dark filled dots) and controls (white dots). Transcriptome data further separated samples from IA patients by aneurysm size, forming groups of large IAs (>8 mm, with one exception - see asterisk) and small IAs (<5 mm). (B) Multidimensional scaling (MDS) of transcriptome data further reduces dimensionality and mirrors the PCA results, also showing separation of IA (dark filled dots) and control (white dots) samples. (C) Hierarchical clustering using only the 82-transcript IA- associated signature also demonstrates separation of IA and control groups. Four aneurysms samples on the right were more distinct than others, while the rest of the samples segregated into three main clusters, two containing all control and one containing all IA (with one exception). Overall, 91% of the samples were grouped with their respective group.

[0018] Figure 4. The 4 most regulated networks. Networks were derived from IP A of differentially expressed transcripts (p<0.05) in neutrophils from IA patients and controls. Fold-change is represented by intensity of grey shading. Non-differentially expressed transcripts with known interactions have no grey color. (A) This network (p-score=41) shows regulation of transcripts with increased expression by an ERK1/2 and API . IL8, regulates transcripts with lower expression in samples from patients with IAs. (B) This network (p- score=30) shows regulation of transcripts by UBC. (C) This network (p-score=30) shows two nodes of regulation at AKT and VEGF. (D) This network (p-score=23) shows regulation of transcripts with lower expression in IA samples by IFNG.

[0019] Figure 5. Replication study in a cohort of 10 new patients (5 with IAs). (A)

Principal component analysis performed using the 82 IA-associated transcripts shows separation of IA patients (dark filled dots) from controls (white dots) in this unmatched cohort. (B) Hierarchical cluster analysis demonstrates separation of the IA and control samples, with the exception of one IA sample that was grouped with controls.

[0020] Figure 6. White blood cell populations in the IA and control groups. There was no significant difference in white blood cell count or leukocyte ratios between patients with IAs (n=l 1) and controls (n=7, no data were available for 4 of the controls). (A)

Complete blood count data recorded within 3 months of blood collection showed no significant difference between-groups in the concentrations of leukocytes, erythrocytes, platelets, neutrophils, lymphocytes, or monocytes (p>0.05, Student's t-test). (B) There was also no significant difference in the percentage (%) per 100 leukocytes for neutrophils, lymphocytes, monocytes, eosinophils, and basophils between patients with and without IA (p>0.05, Student's t-test). (Data points=average values, error bar s= standard error).

[0021] Figure 7: Neutrophil RNA expression differences between patients with

IA and IA-free controls and feature selection for classification model creation. (A) Transcriptome profiling demonstrated 95 differently expressed transcripts (q-value<0.05) between patients with intracranial aneurysm (IA) and controls. Of these, 26 had an FDR<0.05 and an absolute fold change >1.5 (in grey). (B) Principal component analysis (PCA) using these 26 transcripts demonstrated general separation between patients with IA (dark filled dots - 60% circled labeled "IA") and controls (white dots - 80% circled labeled "control"). (C) Estimation of model performance during LOO cross-validation in the training cohort demonstrated that most models performed with an accuracy of 0.50-0.73. Overall Diagonal Linear Discriminant Analysis (DLDA) had the highest combination of sensitivity, specificity, and accuracy (0.67, 0.80, 0.73). (D) Receiver Operator Characteristic (ROC) analysis using classifications in the training dataset demonstrated that the models had Area Under the Curve (AUC) of 0.54 (SVM) to 0.72 (DLDA).

[0022] Figure 8: Performance of the 4 classification models during model training and testing. (A) Principal component analysis using the 26 transcripts demonstrated general separation between patients with IA (dark filled dots - 100% circled labeled "IA") and controls (white dots - 80% circled labeled "control"). (B) Validation of the classification models in an independent testing cohort of patients demonstrated that DLDA had the best performance, with sensitivity, specificity, and accuracy of 0.80, 1.0, and 0.90, respectively. (C) ROC analysis in the testing cohort demonstrated that DLDA also had the best AUC (0.80).

[0023] Figure 9: Assessment of the model performance by LOO cross-validation over all data, and positive predictive value (PPV) Negative Predictive Value (NPV). (A) Estimation of model performance showed that the models performed with an accuracy of 0.63-0.80. DLDA had the highest combination of sensitivity, specificity, and accuracy (0.65, 0.95, 0.80). (B) ROC analysis demonstrated that the models had AUC of 0.68 (Nearest Shrunken Centroids (NSC)) to 0.84 (DLDA). (C) The positive predictive value of all models across all possible prevalence. The best performing model (DLDA) had the highest PPV, and Support Vector Machines (SVM) demonstrated poorest PPV. (D) The DLDA models also had the best NPV, but only slightly better than that of the cosine NN, NSC, and SVM.

[0024] Figure 10: Validation of RNA-Seq data for 7 transcripts by Quantitative

Polymerase Chain Reaction (qPCR). Six of seven differentially expressed transcripts in patients with and without IA were also differentially expressed in neutrophils in the qPCR in an independent cohort.

DETAILED DESCRIPTION

[0025] Unless defined otherwise herein, all technical and scientific terms used in this disclosure have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains.

[0026] Every numerical range given throughout this specification includes its upper and lower values, as well as every narrower numerical range that falls within it, as if such narrower numerical ranges were all expressly written herein.

[0027] The present disclosure provides methods for the diagnosis and/or prediction and/or intervention and/or treatment of aneurysms. In embodiments the disclosure relates to intracranial aneurysms. The present disclosure also provides methods for the determination of aneurysm size and location by medical imaging following determination of aneurysm presence. The present disclosure also provides methods for determining the risk of subarachnoid hemorrhage (SAH) and biomarkers useful for predicting such risk, since the intracranial aneurysms are responsible for as much as 90% of all non-traumatic SAH.

[0028] The present disclosure provides a method for determining the presence or absence of an intracranial aneurysm in a subject suspected of having an intracranial aneurysm, or at risk for developing an intracranial aneurysm, which method comprises analyzing a biological sample from the subject for differential expression of a biomarker as described herein. A subject suspected of having an intracranial aneurysm or at risk for an intracranial aneurysm is a subject who may be experiencing a headache (such as a pain above and behind an eye), a dilated pupil, a significant vision change, double vision, a drooping eyelid, numbness on a side, weakness on a side, paralysis on a side, difficulty with memory or speech, a seizure. In an embodiment, the side is a facial side. Additionally, a subject suspected of having an intracranial aneurysm or at risk for an intracranial aneurysm may have no symptoms at all. A subject may be at risk for an intracranial aneurysm due any of the following: old age, female gender, smoking, family history of aneurysm, hypertension, hyperlipidemia, and/or heart disease. In this application, the term, biomarker, is used to mean RNA or protein translated from said RNA. The RNA that is determined is typically coding RNA, such as a section of mRNA that codes for a protein described herein, or it may be non- coding RNA. As will be recognized by those skilled in the art, the presence, absence, and/or amount of RNA can be determined from, for example, DNA amplified from the RNA template, as well as by other methods that are described herein.

[0029] In developing the present disclosure we asked if circulating neutrophils carry transcriptional signatures of unruptured IA. Our rationale was twofold. Firstly, aneurysmal lesions are known to be associated with persistent vascular wall inflammation, and are in direct contact with circulating immune cells, including neutrophils, the most abundant white blood cell. Secondly, neutrophils, while generally considered non-specific in their function, have been shown to alter their transcriptomes in diseases characterized by inflammation - even short exposure to specific physiologic contexts can subtly alter their transcriptional programs. Indeed, RNA expression differences in circulating neutrophils have been demonstrated in idiopathic arthritis, sepsis, lung cancer, and the response to xenografts.

Analysis of whole blood transcriptomes has also demonstrated that expression differences correlate with localized vascular diseases, including atherosclerosis, thoracic aortic aneurysm, coronary artery disease, and arteriovenous malformations.

[0030] Circulating neutrophils, being in continuous interaction with the aneurysm tissue, could also carry a signature of IA in their transcriptomes. Therefore, in at least the first Example of this disclosure, we asked if patients with IA present with different neutrophil RNA expression profiles compared to subjects without aneurysms. To this end, we matched two cohorts of patients with and without IAs by demographics and comorbidities, and performed next-generation RNA sequencing and an array of bioinformatics analyses to identify and characterize IA-associated expression differences in circulating neutrophils. Furthermore, to determine if biomarker differences revealed by RNA sequencing could be detected using inexpensive methods in a broader population, we performed a corroboration using RT-qPCR in a new, unmatched cohort of patients.

[0031] In embodiments, the disclosure provides a signature of biomarkers.

"Signature" as used herein means a combination of informative biomarkers, but is not meant to exclude other biomarkers from combinations of markers that can be used in methods of this disclosure. For instance, an 82 biomarker signature is described in Example 1 below (see, Example 1, Table 2), but the signature can comprise alternative and/or additional biomarkers, 16 of which are described in Example 2. In particular, Example 2 expands on Example 1 to provide a combination of 26 biomarkers, which includes a set of 10 markers from the 82 biomarkers that are provided in Example 1, Table 2. The biomarkers described in Example 2 were selected for classification model training, and also constitute a biomarker signature as described herein.

[0032] Example 1, Table 2 includes Entrez Gene ID numbers for each of the 82 genes from which the RNA is transcribed, and GenBank Accession numbers for each transcript. Example 2, Table 3 provides 26 biomarkers, 10 of which overlap with the biomarkers described in Example 1, and includes GenBank Accession numbers for each transcript. The additional 16 biomarkers are in bold font in Example 2, Table 3. "Transcript" as used herein includes a cDNA sequence of an mRNA.

[0033] The sequences of each RNA (or a corresponding cDNA) can be determined by those skilled in the art using the Gene ID numbers and accession numbers provided herein. All sequences for all GenBank accession numbers described herein are incorporated herein by reference as they exist in the GenBank database as of the priority date of this application or patent. The disclosure includes all of the polynucleotide and amino acid sequences in these database entries, the mRNA equivalent of any cDNA entry, the cDNA equivalent of any RNA entry, and includes all isoforms and splice variants, if any, of such sequences.

[0034] It will be recognized form this disclosure post hoc power estimation and independent corroboration of expression differences indicates that the biomarker signatures described herein are consistently present in patients with IA. Thus, without intending to be bound by any particular theory, it is believed that this is the first description of biomarker signatures for unruptured IA, and in particular for biomarkers that are differentially expressed in in circulating neutrophils.

[0035] The disclosure includes determining the presence, absence, expression level and/or amount of at least one of the biomarkers described herein. Accordingly the disclosure includes analyzing 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, 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, or [0036] In certain embodiments, the invention comprises a method of determining the presence, absence, expression level and/or amount of at least one biomarker selected from the group consisting of C21orfl5, CYP1B 1, FLT3, XKR3, SLC12A7, PAM, GPC4, FBN1, IL-8, GBP 5, ETV7, MFSD9, SERPING1, TCL1A and CARD17. In certain embodiments, the presence, absence, expression level and/or amount of 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 or 15 of these biomarkers are analyzed. In certain embodiments disclosure comprises a method of determining the presence, absence, expression level and/or amount of at least one biomarker selected from the group consisting of IL8, GBP 5, ETV7, MFSD9, SERPING1, TCL1A, CARD 17, PAM, XKR3, CYP1B1 and FBN1. IL8, GBP 5, ETV7, MFSD9,

SERPING1, TCL1A, CARD 17, PAM, XKR3, CYP1B1 are the top 10 transcripts from a nearest shrunken centroids analysis as described in Example 1. FBN1, XKR3, ETV7, and MFSD9 can be used in a regression model to predict probability of aneurysm presence also as described in Example 1.

[0037] As described in Example 2, in embodiments, the invention comprises a method of determining the presence, absence, expression level and/or amount of at least one biomarker selected from the group consisting of PVRL2, CYP1B 1, CD 177, PDE9A,

ARMC12, OLAH, TGS1, CD163, LOC100506229, OCLN, SEMA6B, ADTRP, VWA8, MTR R2L10, HOXB2, EPCAM, IL18R1, IGSF23, PTGES, G0S2, FCRL5, Clorf226, UTS2, HBG2, CYP26B 1, C1QL1, and combinations thereof. In certain embodiments, the invention includes determining the presence, absence, expression level and/or amount of at least one biomarker that includes at least one of PVRL2, PDE9A, TGS1, LOCI 00506229, OCLN, SEMA6B, MTRNR2L10, HOXB2, EPCAM, IL18R1, IGSF23, PTGES, UTS2, HBG2, CYP26B1, and C1QL1.

[0038] In embodiments, the disclosure comprises determining all or any combination of the following markers are upregulated, relative to a control: PVRL2 (NM_002856.2); CYP1B 1 (NM_000104.3); CD177 (NM_020406.3); PDE9A (NM_002606.2); ARMC12 (NM_145028.4); OLAH (NM_018324.2); TGS1 (NM_024831.7); CD163 (NM_004244.5); LOC100506229 (NR_039975.1); OCLN (NM_002538.3); SEMA6B (NM_032108.3);

ADTRP (NM_001143948.1); VWA8 (NM_015058.1); MTRNR2L1 (NM_001190708.1); HOXB2 (NM_002145.3); EPCAM (NM_002354.2); and IL18R1 (NM_003855.3). In an embodiment, the disclosure comprises determining that any one or any combination of the following markers are downregulated, relative to a control: IGSF23, (NM_001205280.1); PTGES, (NM_004878.4); G0S2, (NM_015714.3); FCRL5, (NM_031281.2); Clorf226, (NM_001135240.1); UTS2, (NM_021995.2); HBG2, (NM_000184.2); CYP26B 1, (NM_019885.3); and C1QL1, (NM_006688.4). In embodiments, the disclosure comprises determining that all of the markers in this paragraph are upregulated and downregulated, respectively, relative to a control.

[0039] The invention is generally suitable for use with any biological sample obtained from an individual. The biological sample can be tested directly, or it can be subjected to a processing step to isolate, amplify or purify components of the sample before testing. In certain embodiments the biological sample comprises a liquid biological sample, such as whole blood. If desired the whole blood can be processed prior to testing, such as by separating certain blood components and/or cell types for testing. In certain embodiments, the sample comprises immune cells, such as neutrophils. In certain approaches the neutrophils can be separated from other blood components, such as peripheral blood mononuclear cells and plasma. In embodiments the sample can comprise neutrophils and erythrocytes. In embodiments, the sample comprises neutrophils with eosinophils, basophils, and mast cells, or any combination thereof. In embodiments, the erythrocytes can be lysed before testing the neutrophils. Neutrophils can be isolated from the biological sample by, for example, centrifugation. Neutrophils can be identified if desired by determining the CD66b+ biomarker, such as by flow cytometry. In some embodiments the biological sample tested does not comprise CD 14+ cells as determined by flow cytometry.

[0040] Determining the presence, absence, expression levels and/or amount of the biomarkers described herein can be performed using any suitable approaches. In one approach wherein the biomarker is a protein, protein is determined, such as by using immunological -based approaches, such as any form of ELISA assays. In embodiments, RNA is detected and/or quantitated using any suitable approach including RNA sequencing. In certain implementations PCR amplifications are used, and can include separating RNA, such as total RNA or mRNA from a sample, and making and measuring cDNA from the RNA, and/or by using quantitative PCR approaches, including real-time PCR, such as quantitative reverse transcription PCR (RT-qPCR). In embodiments, determining biomarkers comprises RNA sequencing, SAGE (serial analysis of gene expression), hybridization-based techniques including but not limited to those performed using RNA microarrays, DNA microarrays, or tiling arrays, and in situ hybridizations, northern blotting, and capillary electrophoresis.

[0041] "Upregulated" and its various forms used herein means that expression of a particular mRNA is increased relative to a control. "Downregulated" and its various forms as used herein means expression of a particular mRNA is decreased relative to a control. Non- limiting examples of degrees of upregulation and downregulation are provided in the tables, figures, and examples of this disclosure.

[0042] Quantitative or qualitative determinations of the amount of each biomarker in a sample according to the present disclosure, including but not limited to whether expression of any particular biomarker is upregulated or downregulated, can be measured using any suitable technique and compared to any suitable reference, including but not necessarily limited to an established normal range, a standardized curve, positive, negative, or matched controls, etc. In embodiments, the amount of biomarker is compared to a suitable control value, such as a value obtained from determining the amount of the same biomarker in a reference, wherein the reference comprises one or more individuals who have been determined to be normal and as such have no detectable intracranial aneurysm, or one or more individuals who each have an un-ruptured intracranial aneurysm, or from one or more individuals who each have a ruptured intracranial aneurysm. In embodiments, determining that one or more biomarkers as described herein is the same or similar to a control, such as a normal control, can indicate the individual does not have an aneurysm. In embodiments, determining a difference relative to a control comprises determining a statistically significant difference. In embodiments, a p value is determined. In embodiments, determining a p<0.05 is a statistically significant difference. In embodiments, differences between biomarkers that qualify as distinguishing criteria are those differences that are statistically significant.

[0043] In certain embodiments, a control comprises biomarkers from an individual or individuals who have been determined to have an intracranial aneurism, which may be an aneurism of a certain size, or of a certain size range. With respect to size, it is the most widely used metric for assessing intracranial aneurysm (IA) rupture risk. This metric is typically measured as the length of the largest dimension of the aneurysm on medical imaging and has been adopted from longitudinal prospective studies that found that larger aneurysms are more likely to rupture. Specifically, the 2012 Unruptured Cerebral Aneurysm Study (UCAS) that included 5720 patients who had cerebral aneurysms found that the risk of rupture increased with increasing size of the aneurysm (UCAS Japan Investigators et al. EJM 2012). The risk of rupture for aneurysms that were smaller than 5 mm was less than those that were 5 mm or larger. This analysis indicates that unruptured aneurysms larger than 5 mm should have a higher priority to be considered for treatment. A recent systematic review of the current literature by Malhotra et al. also support these findings (Malhotra et al. Ann Intern Med 2017). In 25 out of 26 studies in the literature, the annualized rupture rate for aneurysms 3 to 5 mm was 0.5%, but was 1% or greater for aneurysms 5 mm and above. In embodiments, the present disclosure thus relates to predicting the size of an aneurysm based on a determination of biomarkers described herein. In embodiments, a determination that an aneurism is at risk of rupture based on a prediction or estimation of its size is followed by a medical

intervention, as described herein. In embodiments, a determination of a risk of IA rupture is based on predicting or estimating a size of an aneurysm as greater than 5mm, by determining a combination of biomarkers as described herein.

[0044] Differential RNA and protein translated from mRNA, expression profiles can be developed and used in embodiments of the disclosure. In certain aspects the disclosure comprises determining increased expression and decreased expression of biomarkers relative to one or more suitable controls.

[0045] In certain embodiments, the disclosure comprises determining increased expression of at least one of C21orfl5, CYPIBI, FLT3, XKR3, SLC12A7, and PAM, relative to a control. In certain embodiments the disclosure comprises determining decreased expression of at least one of GPC4, FBNl, and IL-8 relative to a control. In certain embodiments, the disclosure comprises determining increased expression of at least one of C21orfl5, CYPIBI, FLT3, XKR3, SLC12A7 and PAM and determining decreased expression of at least one of GPC4, FBNl, and IL-8 relative to a control. In certain embodiments the disclosure comprises determining increased expression of 2, 3, 4, 5 or 6 of C21orfl5, CYPIBI, FLT3, XKR3, SLC12A7 and PAM relative to a control. In certain embodiments the disclosure comprises determining decreased expression of 2 or all three of GPC4, FBNl, and IL-8 relative to a control. In certain embodiments the disclosure comprises determining increased expression of all of C21orfl5, CYPIBI, FLT3, XKR3, SLC12A7 and PAM and determining decreased expression of all GPC4, FBNl, and IL-8 relative to suitable controls. In certain embodiments, the analyzing comprises determining increased expression of at least one of C21orfl5, CYPIB I, FLT3, XKR3, SLC12A7, PAM, and TCLIA relative to a control, and/or determining decreased expression of at least one of GPC4, FBNl, IL-8, GBP 5, ETV7, MFSD9, SERPING1, CARD17 relative to a control.

[0046] In certain embodiments, the disclosure comprises determining increased expression of at least one of PVRL2, CYPIBI, CD177, PDE9A, ARMC12, OLAH, TGSl, CD163, LOC100506229, OCLN, SEMA6B, ADTRP, VWA8, MTRNR2L10, HOXB2, EPCAM, IL18R1, relative to a control. In certain embodiments, the disclosure comprises determining increased expression of at least one of PVRL2, PDE9A, TGSl, LOCI 00506229, OCLN, SEMA6B, MTRNR2L10, HOXB2, EPCAM, and IL18R1, relative to a control. [0047] In certain embodiments the disclosure comprises determining decreased expression of at least one of IGSF23, PTGES, G0S2, FCRL5, Clorf226, UTS2, HBG2, CYP26B 1, and C1QL1, relative to a control. In certain embodiments the disclosure comprises determining decreased expression of at least one of IGSF23, PTGES, UTS2, HBG2, CYP26B1, and C1QL1, relative to a control.

[0048] In embodiments, the disclosure includes a computer-implemented process to determine or otherwise analyze biomarker expression, and/or compare biomarker expression to a control, and includes use of any software program described herein. In embodiments, the disclosure comprises fixing the determination of the biomarkers as disclosed herein in a tangible medium of expression, such as a compact disk, a DVD, or any other form of electronic file. Thus, tangible forms of media comprising a biomarker determination as set forth herein are included in the present disclosure. In embodiments, the invention includes communicating or otherwise transferring the tangible medium comprising the diagnostic determination to a health care provider, such as by electronically transmitting a file containing the determination to the health care provider.

[0049] In certain embodiments, the sample is obtained from an individual who is at risk for developing an aneurysm.

[0050] In certain aspects, determining differences in expression in one or any combination of the biomarkers relative to a control comprises a diagnosis of the presence or, and/or the likely presence of an aneurysm. In embodiments, determining differences in expression in one or any combination of the biomarkers relative to a control aids in a physician's diagnosis of an aneurysm. In certain implementations, determining differences in expression in one or any combination of the biomarkers relative to a control can be followed by a medical intervention, including but not necessarily limited to an angiogram, computed tomography angiography (CTA), digital subtraction angiography (DSA) or Magnetic Resonance Angiography (MRA), and surgical interventions, such as introducing a stent, endovascular clipping, coiling, stenting, Onyx treatment, carotid artery ligation, other occlusive surgeries, bypass or any combination of the forgoing, or any other medical device to inhibit rupture of the aneurysm. Imaging techniques such as computed tomography angiography (CTA), digital subtraction angiography (DSA) or Magnetic Resonance

Angiography (MRA) can be used for confirmation of the presence or absence of an aneurysm. Thus, the invention may comprise a method for identifying a subject in need of confirmation of the presence of intracranial aneurysm(s), its location and geometry by imaging. The invention may further include an assessment of the risk of aneurysm rupture and, optionally, treatment plans.

[0051] In one embodiment, the differential analysis of the biomarker(s) is followed by determining the size, phenotype and location of the intracranial aneurysm by imaging.

Determining aneurysm phenotype (or pathological sub-type) is important for clinical decision making because it can indicate which ones are dangerous (going to rupture) and need treatment or which ones are not dangerous and can be periodically monitored. Aneurysm size is one clinical parameter used to judge the rupture risk of an aneurysm: the UCAS study (UCAS Japan Investigators et al. EJM 2012) demonstrated that aneurysms >5 mm had a greater probability of rupturing. Our data shows the differential RNA expression segregates patients by aneurysm size. (Figure 2B and C).

[0052] The invention also provides a method for determining the risk of subarachnoid hemorrhage (SAH) comprising analyzing a biological a sample for a differential expression of a biomarker as described herein. Subarachnoid hemorrhage (SAH) typically occurs due to bleeding of a ruptured intracranial aneurysm in the subarachnoid space surrounding the Circle of Willis at the base of the brain. The AANS reports that as much as 90 percent of SAHs can be attributed to the rupture of an IA. Hence, the presence of IA is likely the most correlative risk factor for non-traumatic SAH. The only know way to prevent SAH is to identify potential cerebral vascular complications before they occur. Detection of an unruptured IA before it ruptures is the only way to identify SAH early. Once detected, IAs can be treated, which has been demonstrated in the current literature to decrease the risk of future SAH. D' Souza J Neurosurg Anesthesiol 2015: The incidence of SAH in the United States is approximately 30,000 per year. (Brisman et al. NJM 2006). Ruptured IAs account for anywhere from 75% to 85% of non-traumatic SAH.(van Gijn and Rinkel Brain 2001).

[0053] In certain approaches the disclosure can be used to assess risk of aneurysm rupture or the risk of SAH. Current medical research uses size to delineate aneurysms at risk of rupture. Figure 2B and C and other data presented herein show that expression of neutrophils can separate aneurysm patients by the size of their aneurysm. The analysis in the figures was performed using whole Transcriptome data. Transcriptome data separated aneurysm samples by aneurysm size in the principal component space, forming groups of large IAs and small IAs.

[0054] These biomarkers may also increase or decrease as the aneurysm increases in size. Likewise, some of the biomarkers may increase or decrease dramatically if the aneurysm ruptures. Thus, they could be used to monitor aneurysm size or risk of progression to rupture.

[0055] Biomarker expression differences may also distinguish fusiform aneurysms from berry aneurysms, thin-walled IAs from thick-walled IAs, rupture-prone aneurysms from stable aneurysms etc. Fusiform aneurysms are aneurysms where the entire vessel is dilated. In saccular or berry aneurysms, the aneurysm forms a "berry" in part of the vessel. Fusiform aneurysms have fewer options to treat than saccular ones: they are generally not suitable for surgical clipping and for endovascular coiling. A suitable way to treat fusiform lesions is the recently emerged flow diverter treatment— deploying a densely woven stent mesh along the path of the parent vessel as a new conduit to divert the flow away, relying on the thrombosis (clotting) of the aneurysm and the reconstruction of the parent vessel to happen. Bypass is another treatment option. Irregularly shaped aneurysms are considered more likely to rupture than regularly shaped ones.

[0056] In embodiments the disclosure is used to monitor an individual, and thus multiple analyses of the presence, absence, expression level and/or amount of the biomarker can be obtained prior to, during, and/or subsequent to a medical intervention for confirming the presence of an aneurysm or a medical intervention for treating an aneurysm and/or reducing the risk of aneurysm rupture. This assessment may be periodically performed to monitor the subject. Accordingly, in certain aspects the disclosure comprises analysis of the biomarkers as described herein in a subject in need of aneurysm monitoring. The subject can be in need of monitoring for an aneurysm that was initially detected using any approach of this disclosure, or was detected using any other approach.

[0057] In embodiments, the disclosure comprises performing one or a combination of medical procedures described herein on an individual from whom a sample as described herein indicates the presence of an aneurysm. In embodiments, the sample indicates the presence of an IA that is at risk of rupture, and/or exceeds a size value, such as a threshold size value, as described herein. In embodiments, the disclosure thus comprises performing a medical procedure on an individual based at least in part on a result obtained by performing a method described herein. The disclosure accordingly includes selecting an individual to receive the medical procedure based at least in part on a result that includes a value for a combination of biomarkers, as described herein. In embodiments, the medical procedure comprises imaging of an area of the brain of the individual to confirm the presence, absence, location, size, ruptured or un-ruptured status, etc., of one or more IAs. In embodiments the medical imaging procedure comprises any of X-Ray, magnetic resonance imaging (MRI), magnetic resonance angiography (MRA), computed tomography (CT), or digital subtraction angiography (DSA). In embodiments, the medical procedure comprises a surgical procedure. In embodiments, the surgical procedure comprises clipping, which is well known in the art as a surgical procedure for treating aneurisms, and involves introducing a clip at the base of the aneurysm to inhibit blood flow into the location of the aneurysm. In another embodiment, the surgical procedure comprises endovascular coiling, which is also a well know procedure for treating aneurisms, and involves inserting a coil into the aneurism to inhibit blood from entering it. In embodiments, the medical procedure comprises administering one or more pharmaceutical agents to the individual. In embodiments, the pharmaceutical agent comprises an agent that promotes blood clotting, such as an antifibrinolytic agent, such as aminocaproic acid or tranexamic acid.

[0058] In embodiments, the disclosure further comprises recommending a treatment protocol to an individual based at least in part on a diagnosis made by determining one or more biomarkers as described herein. The medical intervention or recommended treatment protocol may optionally comprise supplemental treatments, such as high blood pressure control and/or lifestyle modifications such as smoking cessation and/or weight loss.

[0059] The following Examples are meant to illustrate but not limit the invention. Example 1

[0060] In this Example, transcriptome profiling identified 258 differentially expressed transcripts in patients with and without IAs. Expression differences were consistent with peripheral neutrophil activation. An IA-associated RNA expression signature was identified in 82 transcripts (p<0.05, fold-change >2). This signature was able to separate patients with and without IAs on hierarchical clustering. Furthermore, in an independent, unpaired, replication cohort of patients with IAs (n=5) and controls (n=5), the 82 transcripts separated 9 of 10 patients into their respective groups.

[0061] In more detail, in this Example, we investigated whether neutrophils have different RNA expression profiles in patients with IAs compared to patients without IAs. We recruited patients with and without aneurysms (confirmed on angiography) and paired them based on demographics and comorbidities. Next-generation RNA sequencing of circulating neutrophils was performed to identify an IA-associated expression signature in their transcriptomes. We further assessed if the IA-associated expression signature could distinguish patients with and without IA in a heterogeneous independent cohort of patients. Gene ontology analysis and physiological pathway modeling were used to determine the biological function of differentially expressed transcripts in IA.

[0062] The following materials and methods were used to obtain the results described in this Example.

[0063] Clinical study

[0064] The study described in this Example was approved by an institutional review board. Methods were carried out in accordance with the approved protocol. Written informed consent was obtained from all subjects. Peripheral blood samples were collected from patients undergoing cerebral digital subtraction angiography (DSA): 35 patients had a positive IA diagnosis and 42 had a negative IA diagnosis (controls). Positive or negative IA diagnosis was confirmed by imaging, and patient medical records were collected for pairing patients with IAs to controls. Additionally, each patient's complete blood count, which was taken within 3 months of blood collection, was recorded.

[0065] Patients undergoing cerebral digital subtraction angiography (DSA) with positive and negative intracranial aneurysm (IA) diagnoses were enrolled in this study.

Reasons for the patients to receive DSA included confirmation of findings from noninvasive imaging of the presence of IAs, vascular malformations, or carotid stenosis or follow-up noninvasive imaging of previously detected IAs. All consenting patients were older than 18 years, were English speaking, and had not received previous treatment for IA. To ensure that differences in the circulating neutrophils were not influenced by inherent inflammatory conditions, we excluded patients who potentially had altered leukocyte transcriptomes; this included patients who were pregnant, had recently undergone invasive surgery, were undergoing chemotherapy, had a body temperature above 37.78°C (100°F), had received solid organ transplants, had autoimmune diseases, and those who were taking prednisone or any other immunomodulating drugs. Furthermore, the included patients did not have any other known cerebrovascular malformations or extracranial aneurysms, including abdominal aortic aneurysms.

[0066] Sample preparation

[0067] Sixteen mL of blood was drawn from the access catheter in the femoral artery and transferred into two 8 mL, citrated, cell preparation tubes (BD, Franklin Lakes, NJ).

Neutrophils were isolated within 1 hour of peripheral blood collection, according to known techniques. Cell preparation tubes were centrifuged at 1,700 ^χ g for 25 minutes to separate erythrocytes and neutrophils from mononuclear cells and plasma in the peripheral blood samples. Erythrocytes and neutrophils were collected into a 3 mL syringe and placed into an erythrocyte lysis buffer that was made in-house. After all erythrocytes were lysed, the neutrophils were isolated by centrifugation at 400 x g for 10 min and disrupted and stored in TRIzol reagent (Life Technologies, Carlsbad, CA) at -80 °C until further processing.

Neutrophils isolated in this fashion are more than 98% CD66b+ by flow cytometry and contain no contaminating CD 14+ monocytes

[0068] Total neutrophil RNA was extracted using TRIzol, according to the manufacturer's instructions. Trace DNA was removed by DNase I (Life Technologies, Carlsbad, CA) treatment. The RNA was purified using the RNeasy MinElute Cleanup Kit (Qiagen, Venlo, Limburg, Netherlands) and suspended in RNase-free water. After RNA isolation, the purity and concentration of RNA in each sample was measured by absorbance at 260 nm on a NanoDrop 2000 (Thermo Scientific, Waltham, MA), and 200-400 ng of RNA was sent to our university's Next-Generation Sequencing and Expression Analysis Core facility for further quality control. Precise RNA concentration was measured at the core facility via the Quant-iT RiboGreen Assay (Invitrogen, Carlsbad, CA) with a TBS-380 Fluorometer (Promega, Madison, WI), whereas the quality of the RNA samples was measured with an Agilent 2100 BioAnalyzer RNA 6000 Pico Chip (Agilent, Las Vegas, NV). RNA samples with 260/280 > 1.9 and an RNA integrity number (RIN) > 6.0 were considered for RNA sequencing.

[0069] Cohort creation

[0070] Before sequencing, samples from IA patients and control subjects were paired by demographics and comorbidities to control for confounding variables. First, samples that did not have acceptable RNA quality for sequencing were excluded. Next, each patient in the IA group was paired with a control subject by factors that have been reported in the literature to correlate with IA. These included (in order of decreasing importance) age, sex, smoking status (yes or no), presence of hypertension, presence of hyperlipidemia, and presence of heart disease. Matching criteria also included stroke history, presence of diabetes mellitus, and presence of osteoarthritis, when possible. With the exception of age, the factors used for matching were quantified as binary data points. The clinical factors were retrieved from the patients' medical records via a Patient Medical History form administered prior to imaging. Since this medical record contained self-reported information, the presence of each comorbidity was corroborated with each patients' reported list of medications (e.g.

hypertension with lisinopril, hyperlipidemia with simvastatin, heart disease with metoprolol, stroke history with clopidogrel, diabetes mellitus with metformin, and osteoarthritis with NSAIDs/tramadol). We were able corroborate 84% of the clinical data points for patients' comorbidities through their medication history.

[0071] After performing the original experiments to identify an IA-associated neutrophil expression signature, we used the same clinical protocol to recruit an additional 5 patients with IAs and 5 IA-free controls into a small replication cohort (n=10) to test whether the IA-associated signature could separate IA patients from controls in the second cohort. Blood samples and RNA were handled in the same manner as those in the original cohort, and the same RNA sequencing and data analysis protocols were followed. However, prior to sequencing, we did not control for demographics and comorbidities to obtain a more heterogeneous cohort.

[0072] RNA sequencing

[0073] RNA libraries for these samples were constructed at our university's Next-

Generation Sequencing and Expression Analysis Core facility using the TruSeq RNA Library Preparation Kit (Illumina, San Diego, CA). All samples were subjected to 50-cycle, single- read sequencing in the HiSeq2500 (Illumina) and were demultiplexed using Bcl2Fastq v2.17.1.14 (Illumina). Gene expression analysis was completed using the Tuxedo Suite. Short RNA fragment data were compiled in FASTQ format and aligned to the human reference genome (human genome 19 [hgl9]) using TopHat v2.1.13. Gene expression levels were calculated using fragments per kilobase of transcript per million mapped reads (FPKM) normalization in CuffLinks v2.2.1 RNA sequencing data files and processed transcript expression have been made available at NCBI's GEO (accession no. GSE106520). To evaluate the quality of RNA sequencing, we performed quality control analysis using both FASTQC before alignment and MultiQC after alignment.

[0074] Differential expression analysis

[0075] Differential gene expression analysis was performed in CuffDiff v2.2.1 and visualized in the CummeRbund v2.7.1 package in R. We used CuffDiff v2.2.1 (Trapnell Laboratory), which compared the log ratio of FPKM values in the IA and control groups against the log ratio of FPKM values of the IA group, and computed a test statistic. The test statistic was calculated using the negative binomial distribution to model the variance of each sample and the square root of the Jensen-Shannon divergence to assess differences in relative abundance. The change in Jensen- Shannon divergence was then assigned a p-value, according to a known approach.

[0076] Transcripts were considered significantly differentially expressed at p<0.05.

We defined an IA-associated expression signature as those significant transcripts that also had an absolute fold-change >2. Post hoc power estimation was performed following Hart et al. (Hart SN, Therneau TM, Zhang Y, Poland GA, Kocher JP (2013) Calculating sample size estimates for RNA sequencing data. J Comput Biol 20: 970-978) with a=0.05, an average coefficient of variation of 0.404 (calculated from FPKMs), and counts per million mapped reads of 38. Multiple testing correction was performed by using the Benjamini-Hochberg method (Benjamini Y, Hochberg Y (1995) Controlling the False Discovery Rate: A Pratical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society Series B (Methodological) 57: 289-300), and q-values were reported for each transcript.

[0077] Verification by RT-qPCR

[0078] To verify expression differences measured by RNA sequencing, quantitative reverse transcription polymerase chain reaction (RT-qPCR) was performed. We verified expression difference of 5 differentially expressed transcripts (CD 177, SERPINGl, GBP5, IL8, NAAA) in order to conserve RNA material. These 5 transcripts were chosen because they were among the most prominently differentially expressed transcripts, i.e., they were highly abundant (FPKM>10) and significantly differentially expressed (p<0.05) with an absolute fold-change >1.5. For each transcript, oligonucleotide primers were designed with a -60 °C melting temperature and a length of 15-25 nucleotides to produce PCR products with lengths of 50-200 base pairs using Primer3 software and Primer BLAST (NCBI, Bethesda, MD). The replication efficiency of each primer set was tested by performing qPCR on serial dilutions of cDNA samples (primer sequences, annealing temperatures, efficiencies, and product lengths are shown in Example 1, SI Table).

[0079] For reverse transcription, first-strand cDNA was generated from total RNA using Omni Script Reverse Transcriptase kit (Qiagen, Venlo, Limburg, Netherlands) according to the manufacturer's instructions. Quantitative PCR was run with 10 ng of cDNA in 25 μ reactions in triplicate in the Bio-Rad CFX Connect system (Bio-Rad, Hercules, CA) using ABI SYBR Green Master Mix (Applied Biosystems, Foster City, CA) and gene- specific primers at a concentration of 0.02 μΜ each. The temperature profile consisted of an initial step of 95°C for 10 minutes, followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 minute, and then a final melting curve analysis from 60°C to 95°C for 20 minutes. Gene-specific amplification was demonstrated by a single peak using the Bio-Rad dissociation melt curve. As previously described (Jiang K, Sun X, Chen Y, Shen Y, Jarvis JN (2015) RNA sequencing from human neutrophils reveals distinct transcriptional differences associated with chronic inflammatory states. BMC Med Genomics 8: 55), GAPDH expression was used for normalization, and fold-changes between groups were calculated using the 2 ^{" Ct} method.

[0080] Dimensionality reduction

[0081] We performed dimensionality reduction by unsupervised principal component analysis (PCA) and multidimensional scaling (MDS) using the transcriptomes of each sample in the CummeRbund and prcomp packages in R Bioconductor under the default settings. For hierarchical clustering, we used the hclust package in R. Dendrograms were created using Ward linkage from z-score normalized transcript levels.

[0082] Bioinformatics

[0083] We performed gene set enrichment analysis using the open-source software

GO: :TermFinder (Stanford University School of Medicine, Stanford, CA). This tool determined whether any gene ontology terms annotated two lists of genes (i.e., genes with higher expression in samples from patients with IAs than those without IA and genes with lower expression in samples from patients with IAs than those without IA) greater than what would be expected by chance. Significantly enriched ontologies were reported if the Q-Value was <0.05, based on significance testing using the hypergeometric distribution.

[0084] Networks of potential interactions were generated using Ingenuity Pathway

Analysis (IP A) software (Qiagen Inc., www.qiagenbioinformatics.com/products/ingenuity pathway-analysis). For IP A, each gene identifier was mapped to its corresponding gene object in the Ingenuity Knowledge Base and overlaid onto a molecular network derived from information accumulated in the Knowledge Base. Gene networks were algorithmically generated based on their "connectivity" derived from known interactions between the products of these genes. Networks were considered significant if their p-scores were >21.

[0085] The following results were obtained using the foregoing materials and methods.

[0086] Study participants

[0087] During the 6-month study period, we collected 77 blood samples (35 from patients with IA, 42 from control subjects) as well as angiographic images and medical records data from individuals undergoing cerebral DSA. Of the blood samples collected, 37 (16 from IA patients, 21 from controls) had a sufficient quality of neutrophil RNA for sequencing. Pairing on the basis of demographics and comorbidities resulted in a final cohort of 22 individuals, including 11 IA patients and 11 IA-free controls (Example 1, Table 1). These samples were of sufficient quality and had an average 260/280 of 2.02 and an average RIN of 7.04 (Example 1, S2 Table). Patients with IAs had aneurysms ranging in size from 1.5-19 mm, and included 3 individuals with multiple IAs (Example 1, S3 Table). There was no statistical difference in age (p>0.05, Student's t-test), and other factors (p>0.05, χ ² test) (Example 1, Table 1) as well as white blood cell populations between the two groups (p>0.05, Student's t-test) (Figure 6).

[0088] Example 1, Table 1. Clinical characteristics

Patients with IA Patients without

(n=ll) IA (n=ll) P-Value

Age (years) (MeaniSE) 66.91±2.84 64.73±4.22 0.67

Age (years) [Median (Ql/Q3)l 67 (60.5/72) 70 (60/71.5)

Sex

Female 63.64% 54.55% 0.66

Current Smoker

Yes 18.18% 18.18% 1.00

Comorbidities

Hypertension 63.64% 81.82% 0.34

Hyperlipidemia 45.45% 54.55% 0.67

Heart disease 18.18% 27.27% 0.61

Stroke history 0.00% 9.09% 0.31

Diabetes mellitus 18.18% 36.36% 0.34

Osteoarthritis 27.27% 27.27% 1.00

* We controlled for demographics and comorbidities so no factor was significantly higher in patients with IA or without IA (confirmed on imaging). There is no significant difference in age (p>0.05 by Student's t-test) sex, smoking history, and comorbidities (χ ² >0.05, chi-squared test) between the two groups. (IA=intracranial aneurysm, SE=standard error, Q=quartile)

[0089] Neutrophils have an IA-associated RNA expression signature

[0090] We performed RNA sequencing to identify differentially expressed neutrophil transcripts between 11 patients with IA and 11 paired controls. The sequencing had an average of 52.05 million sequences per sample and a 96.3% read mapping rate (% aligned) (Example 1, S4 Table). The volcano plot in Fig 1 shows neutrophil expression differences between IA patients and controls in terms of average fold-change in expression and significance level. From 13,377 transcripts with testable expression differences, we identified 258 transcripts that were significantly differentially expressed (p<0.05) between the two groups. We defined an IA-associated RNA expression signature as significant transcripts that were increased or decreased by a factor of 2 or more. From the 258 transcripts, 82 met these criteria and are shown by the shaded circles in Fig 1 and detailed in Example 1, Table 2. Post hoc power analysis estimated that a power of 0.8 was achieved in detecting expression differences of at least 1.68 fold at a=0.05. Therefore, our statistical criteria of p<0.05 and absolute fold-change>2 had power >0.8 in detecting this signature.

[0091] Example 1, Table 2. 82-transcript intracranial aneurysm-associated gene expression profile*

Gene ID Accession No.

Transcript Log ₂ (F-C) P-Value Q-Value

MAOA 4128 M69226.1 5.56 0.03455 0.9999

C21orfl5 54055 AY040090.1 2.38 0.00005 0.0836063 *

CYP1B1 1545 NM_000104.3 2.02 0.00005 0.0836063 *

ARMC12 221481 NM_145028.4 1.95 0.0006 0.4459

CD177 57126 NM_020406.3 1.81 0.001 0.585244

OLAH 55301 NM_018324.2 1.79 0.0057 0.9999

CYP1B1-AS1 285154 NR_027252.1 1.73 0.004 0.9999

FLT3 2322 NM_004119.2 1.63 0.00005 0.0836063 *

CD 163 9332 DQ058615.1 1.62 0.0005 0.393441

KCNMA1 3778 NM_001014797.2 1.61 0.050 0.9999

DACT1 51339 NM_016651.5 1.60 0.0045 0.9999

FAM90A1 55138 NM_018088.3 1.58 0.0059 0.9999

SCT 6343 AF244355.1 1.58 0.0232 0.9999

LOCI 00131289 100131289 NR_038929.1 1.54 0.0095 0.9999

NOG 9241 NM_005450.4 1.52 0.013 0.9999

SCAMP5 192683 NM_001178111.1 1.52 0.030 0.9999

PTGDS 5730 NM_000954.5 1.47 0.0052 0.9999

KIR2DS4 3809 NM_012314.5 1.45 0.014 0.9999

CYP4F35P 284233 NR_026756. 1.43 0.0082 0.9999

XKR3 150165 NM_001318251.1 1.41 0.00005 0.0836063 *

RPL39L 116832 NM_052969.2 1.40 0.017 0.9999

CDHR2 54825 NM_001171976.1 1.35 0.013 0.9999

ENHO 375704 NM_198573.2 1.35 0.014 0.9999

SLC12A7 10723 NM_006598.2 1.35 0.00005 0.0836063 *

FLJ27354 400761 NR_033981.1 1.34 0.025 0.9999

DGKH 160851 NM_152910.5 1.33 0.0039 0.9999

SDC3 9672 AF248634.1 1.33 0.028 0.9999

THBS1 7057 NM_003246.3 1.32 0.0019 0.85336

RCVRN 5957 NM_002903.2 1.30 0.016 0.9999 AKR1C1 1645 NM_001353.5 1.30 0.030 0.9999

SCRG1 1 1341 NM_001329597.1 1.28 0.027 0.9999

NRG1 3084 NM_013959.3 1.26 0.030 0.9999

AK5 26289 NM_174858.2 1.24 0.0024 0.9999

ITGA 7 3679 NM_001 144996.1 1.23 0.017 0.9999

PAM 5066 NM_000919.3 1.20 0.00005 0.0836063 *

LYPD2 137797 NM_205545.2 1.19 0.028 0.9999

PRUNE 2 158471 NM_015225.2 1.19 0.0009 0.547241

SLC22A17 51310 NM_020372.3 1.16 0.018 0.9999

ADTRP 84830 NM_001 143948.1 1.16 0.0054 0.9999

ADAMTS1 9510 NM_006988.4 1.15 0.0024 0.9999

ECRP 643332 NR_033909.1 1.15 0.049 0.9999

LOCI 00507387 100507387 NR_038402.1 1.1 1 0.040 0.9999

KLRC2 3822 NM_002260.3 1.1 1 0.034 0.9999

AKR1C3 8644 NM_003739.5 1.09 0.0063 0.9999

SEPT 10 15101 1 NM_144710.4 1.08 0.01 1 0.9999

CYYR1 1 16159 NM_001320768.1 1.08 0.037 0.9999

TCL1A 81 15 NM_021966.2 1.07 0.0024 0.9999

VWF 7450 NM_000552.4 1.06 0.010 0.9999

ONLY 10578 NM_001302758.1 1.06 0.0025 0.9999

C4BPA 722 NM_000715.3 1.05 0.0027 0.9999

LINC00482 284185 NR_038080.1 1.05 0.048 0.9999

KIAA1598 57698 BC022348.1 1.04 0.0092 0.9999

PID1 55022 NM_017933.4 1.03 0.0090 0.9999

SERPINF2 5345 NM_000934.3 1.02 0.027 0.9999

VWA8 23078 NM_015058.1 1.01 0.0057 0.9999

CYP4F2 8529 NM_001082.4 - 1.00 0.039 0.9999

FADS2 9415 NM_004265.3 - 1.01 0.01 1 0.9999

VLDLR 7436 NM_003383.4 - 1.03 0.0005 0.393441

CARD 17 440068 NM_001007232.1 - 1.04 0.016 0.9999

IL8 576 AF043337.1 - 1.04 0.0001 0.148633 *

G0S2 50486 NM_015714.3 - 1.05 0.003 0.9999

FBXW8 26259 NM_153348.2 - 1.08 0.0007 0.468195

MFSD9 84804 NM_032718.4 - 1.08 0.0002 0.26754

CCL23 6368 NM_005064.5 - 1.09 0.019 0.9999

Clorf226 400793 NM OO 1 135240.1 - 1.1 1 0.020 0.9999 GBP5 0.642096

BATF2 0.393441

FCRL5 0.9999

SERPING1 0.334425

B4GALNT3 0.9999

PDCD1LG2 0.9999

FBNl 0.0836063

PRSS21 0.797475

ETV7 0.393441

SEPT4 0.9999

EGR2 0.9999

GBP1P1 0.85336

PSORS1C3 0.9999

HRK 0.9999

NEB 0.9999

GPC4 0.0836063

LOC730441 0.9999

* Significantly differentially expressed transcripts with q-value<0.20 (20% FDR) are marked by "*".

[0092] Multiple hypothesis correction identified 9 transcripts with FDR<0.20;

C21orfl5, CYPIBI, FLT3, XKR3, SLC12A7, PAM, IL8, FBNl, and GPC4. Although this correction effectively reduced the number of significant transcripts, in embodiments, all 82 significant transcripts in the aneurysm-associated signature are analyzed. Individual genes that by themselves might not be significant (i.e., meet strict cutoffs of statistical tests that are not designed to find biologically relevant transcripts) could still play important roles in IA pathophysiology. To avoid missing potentially informative transcripts, we included all 82 transcripts in the IA-associated signature and in the clustering analysis.

[0093] We confirmed differential expression of 5 prominent differentially expressed transcripts (CD 177, NAAA, SERPING1, GBP 5, and IL8) using RT-qPCR. Fig 2

demonstrates that the expression differences between patients with and without IA are of the same direction and of similar magnitudes whether calculated from RNA sequencing or RT- qPCR. There was no statistically significant difference in fold-change of expression measured by the two methods (all p-values>0.05, Student's t-test).

[0094] Neutrophil RNA expression discriminates IA from control groups [0095] To visualize how well neutrophil RNA expression differentiated aneurysm samples from control samples, we performed dimensionality reduction analyses by PCA and MDS using all neutrophil transcriptome data (13,377 transcripts). Fig 3A shows that the IA and control groups separated in the principal component space. Similarly, MDS also showed separation of patients with IAs and control subjects (Fig 3B). We also found that the transcriptome data segregated the patients with IAs by the size of each patient's largest IA, forming two groups on both the PCA and MDS plots: large (>8 mm, with one exception) and small (<5 mm) (Fig 3A and 3B).

[0096] Using the expression signature of 82 transcripts (p<0.05 and absolute fold- change >2), we performed supervised hierarchical clustering to determine whether it could also discriminate patients with IAs from controls. On the dendrogram in Fig 3C, samples from IA and control groups are separated. Four samples from the IA group on the right were more distinct than the others. In the rest of the samples, one control (to the left) was separate, and all other samples segregated into 3 groups. Two groups contained all control samples and 1 group contained all IA samples (with 1 exception). In general, hierarchical clustering congregated 91% of the samples with their respective groups.

[0097] Expression differences are consistent with leukocyte activation

[0098] We performed bioinformatics analyses using gene set enrichment analysis and physiological pathway modeling. Some tightly controlled pathways can be regulated by transcripts that show small but significant changes. We performed bioinformatics analysis using all 258 differentially expressed transcripts (p<0.05) regardless of fold-change. As detailed in Example 1, Table 3, gene ontology analysis revealed that genes with higher neutrophil expression levels in the IA group were involved in defense response, leukocyte activation, stem cell maintenance, maintenance of cell number, cell activation, and stem cell development. Genes with lower expression levels in the IA group were involved in immune response and immune system process (Table 3).

[0099] Example 1, Table 3. Gene ontology (GO) analysis.*

GO ID Term P-value Q-value Annotated Genes

Genes Increased in IA

GO:0006952 Defense Response 3.13E-06 2.36E-03 KLRC2, VNN1, C4BPA, CD300E,

SH2D1B, CD247, ONLY, INHBB, CD1D, KIR2DS4, PRF1, ORM2, STAB1, FCER1A, CD86 GO:0045321 Leukocyte 8.23E-06 6.21E-03 VNN1, CD1D, CD7, PRF1, FCER1A,

Activation SH2D1B, SOX4, CD86, CD247

GO:0019827 Stem Cell 8.44E-06 6.37E-03 NOG, TCL1A, KLF10, SCT, SOX4

Maintenance

GO:0098727 Maintenance of Cell 8.95E-06 6.76E-03 NOG, TCL1A, KLF10, SCT, SOX4

Number

GO:0001775 Cell Activation 1.18E-05 8.92E-03 VNN1, CD7, VWF, SH2D1B, SOX4,

CD247, CD1D, PRF1, FCER1A, CD86

GO:0048864 Stem Cell 1.20E-05 9.10E-03 NOG, DAB2, TCL1A, KLF10, SCT,

Development SOX4

Genes Decreased in I A

GO:0006955 Immune Response 1.17E-07 6.85E-05 AIM2, LILRA4, FCRL5, IL8, MOV10,

CYSLTR2, IFI35, PDCD1LG2, RGS1, CD274, CCL23, DDX60, OLFM4,

GBP1

GO:0002376 Immune System 3.32E-07 1.94E-04 MOV10, CEBPE, SMPD3, IFI35,

Process PDCD1LG2, CCL23, OLFM4, GBP1,

AIM2, FCRL5, LILRA4, IL8, CYSLTR2, RGS1, CD274, DDX60

* Gene set enrichment analysis was performed on significantly differentially expressed genes (p<0.05) in peripheral blood samples obtained from patients with intracranial aneurysms (IA). Significantly enriched ontologies present in genes with higher expression levels included defense response, leukocyte activation, stem cell maintenance, maintenance of cell number, cell activation, and stem cell development. Significantly enriched ontologies present in genes with lower expression levels included immune response and immune system

process. Enriched ontologies from the GO database were considered significant at a false discovery rate- corrected p-value (q-value) <0.05.

[0100] Physiological pathway modeling to identify networks of potential interactions revealed 4 networks with 7 signaling nodes forming hubs within the networks (Fig 4). These hubs included ERK1/2 and API; IL8 (CXCL8), AKT and VEGF; UBC; and IFNG. IPA

indicated that these networks were consistent with activation of cellular movement and cardiovascular system function (network A), lipid metabolism (network B), cell-to-cell signaling and energy production (network C), and organismal injury, cell proliferation, and tissue morphology (network D). These functions are pertinent to neutrophil responses to intravascular perturbations. See Example 1, S5 Table for a list of names and biological functions of the transcripts in these 4 networks.

[0101] Replication study in a new, unpaired population

[0102] To determine whether expression of the IA-associated signature could separate patients with IAs from controls in an independent cohort, we performed a replication study.

To do this, we recruited 10 additional patients (5 with IA, 5 IA-free controls) but did not control for demographics and comorbidities in order to assess the signature's potential for segregating patients in heterogeneous populations (see Example 1, S6 Table for clinical characteristics). Patients with IAs had aneurysms ranging in size from 1.4-7 mm and included one individual with multiple aneurysms (Example 1, S7 Table). From these patients' peripheral blood samples, we isolated neutrophils and extracted neutrophil RNA and performed next-generation RNA sequencing to obtain FPKM levels of the 82 IA-associated transcripts. To visualize how these transcripts could distinguish the IA group from the control group, we performed PCA and hierarchical clustering. With the exception of one IA sample, PCA demonstrated separation of the two groups in the principal component space (Fig 5A). Hierarchical clustering mirrored this result, grouping the IA and control samples separately, with the exception of one IA sample (Fig 5B). Deidentified patient metadata is presented in Example 1, S8 Table.

[0103] It will be recognized by the foregoing description in this Example that we performed transcriptome profiling on circulating neutrophils from paired patients with and without IAs and identified an aneurysm-associated signature of 82 transcripts. These transcripts discriminated patients with and without IA in hierarchical cluster analysis. In a replication study, this signature also distinguished patients with IAs from controls in an unpaired cohort.

[0104] Previous efforts in search of circulating aneurysm biomarkers

[0105] The search for circulating biomarkers for unruptured IAs has spanned more than two decades. A meta-analysis (Hussain S, Barbarite E, Chaudhry NS, Gupta K,

Dellarole A, et al. (2015) Search for Biomarkers of Intracranial Aneurysms: A Systematic Review. World Neurosurg 84: 1473-1483) of IA biomarker publications from 1994-2015 found 5 studies that linked IA presence to specific biomolecules in the blood. These studies found that serum elastase-to-Al AT ratios and LP A, VEGF, MCP-1, IL-Ιβ, TNF-a , and GM- CSF levels were elevated in patients with unruptured aneurysms. However, in the present Example, we did not observe significantly higher mRNA levels for these proteins in neutrophils from patients with IA. This may be because these proteins originate from sources other than neutrophils or may not be sufficiently unique to IA to be identified by our analysis.

[0106] One common trait of the previously-identified potential proteins markers, is that they are ubiquitous, being involved in a wide range of physiological and pathological functions. Thus, in addition to IA, they may also signify various vascular diseases. For example, serum VEGF is also increased during peripheral artery stenosis, plasma MCP-1 is also elevated in thromboembolic hypertension, and LPA is elevated in plasma of patients with vascular dementia. Perhaps for this reason, significant follow-up efforts have not been made towards subsequent biomarker development and validation on the basis of these studies.

[0107] An alternative approach to identifying potential biomarkers is to profile the transcriptome of the circulating blood, which affords screening for multitudes of potential markers and can provide insight into novel disease mechanisms that may be specific to IA. Recently, circulating RNA expression signatures of unruptured IAs were found in microarray studies. In IA patients, Jin et al. (Jin H, Li C, Ge H, Jiang Y, Li Y (2013) Circulating microRNA: a novel potential biomarker for early diagnosis of intracranial aneurysm rupture a case control study. J Transl Med 11 : 296) found 77 differentially expressed plasma microRNAs that were involved in proliferation, apoptosis, molecular activation, transport, and differentiation; Li et al. (Li P, Zhang Q, Wu X, Yang X, Zhang Y, et al. (2014)

Circulating microRNAs serve as novel biological markers for intracranial aneurysms. J Am Heart Assoc 3 : e000972) discovered 119 differentially expressed plasma microRNAs related to inflammatory responses and connective tissue disorders; and Sabatino et al. (Sabatino G, Rigante L, Minella D, Novelli G, Delia Pepa GM, et al. (2013) Transcriptional profile characterization for the identification of peripheral blood biomarkers in patients with cerebral aneurysms. J Biol Regul Homeost Agents 27: 729-738) identified 53 differentially expressed mRNAs from peripheral blood mononuclear cells that were related to increased cell proliferation and apoptosis. These findings indicate that IA is associated with altered expression of a large number of transcripts from various circulating blood components. As indicated, some Examples presented herein were conducted using next-generation RNA sequencing. This latest high-throughput technology affords two key advantages over microarray s used in previous investigations: (1) it offers a larger dynamic range, facilitating detection of expression differences in low-abundance transcripts; and (2) it avoids predetermined probes, allowing examination of novel RNAs (i.e., splice variants, non-coding RNAs, gene isoforms). These capabilities led us to discover a signature of 82 transcripts, containing several uncharacterized and/or non-coding RNAs, which cannot be detected on conventional microarrays. They include C21orfl5, LOC100131289, FLJ27354,

LOC100507387, LINC00482, Clorf226, and LOC730441. To our knowledge, these novel transcripts have not been associated with any other diseases.

[0108] We also designed the present approach to avoid common pitfalls of expression profiling studies. First, to avoid misclassification, we used DSA to confirm that the control subjects did not have aneurysms. Previous studies did not perform such imaging. Second, to find RNA expression differences due to the presence of an IA and not confounding factors, we paired the subjects by demographics and comorbidities. Previous studies typically used healthy subjects or spouses as controls. Third, we performed a replication study in an independent, unpaired cohort to investigate whether the signature can distinguish patients with IA in a general population. These measures helped to increase the likelihood that the discovered signature is associated with IA presence.

[0109] Circulating neutrophils and intracranial aneurysms

[0110] Intracranial aneurysm natural history is characterized by mounting

inflammatory responses and progressive degradation of the aneurysmal wall, starting from initial pro-inflammatory changes in smooth muscle cells that lead to overproduction of matrix metalloproteinases (MMPs). Once the aneurysmal pouch is formed, it harbors a

hemodynamic environment conducive to macrophage and neutrophil infiltration into the wall, which is aided by a local increase of plasma chemokines and cytokines (IL-Ιβ, IL-17, TNF- a) in the lumen. These inflammatory infiltrates massively produce MMPs to further degrade the aneurysm wall and advance its growth and rupture. This is evidenced by gene expression studies of human aneurysmal tissues, which found increased matrix degradation processes, inflammatory processes, and production of inflammatory cytokines and chemoattractant proteins in the IA wall. Furthermore, Yu et al. found that differences in DNA methylation in aneurysmal tissue act to promote inflammatory signaling through the F-KB, JNK-STAT, and ERK/J K pathways (Yu L, Wang J, Wang S, Zhang D, Zhao Y, et al. (2017) DNA Methylation Regulates Gene Expression in Intracranial Aneurysms. World Neurosurg 105: 28-36), uncovering a potential epigenetic underpinning to dysregulated inflammation during IA.

[0111] The role of neutrophils in IA pathophysiology may be complex and is not well understood. Besides secreting MMP-9, activated neutrophils also release NGAL and MPO, which indirectly contribute to extracellular matrix degradation and cytotoxicity, respectively. Increased NGAL in aneurysm tissue modulates the activity of MMP-9, protecting it from degradation and thus aiding aneurysm progression. Increased MPO, an inflammatory enzyme, elicits oxidative stress and pro-inflammatory cell signaling through production of reactive oxygen species. It has been observed that plasma levels of NGAL and MPO are increased in the blood of patients with aneurysms. Furthermore, both of these proteins can have autocrine effects that promote neutrophil activation, which could lead to expression changes observed in our study. Interestingly, we found significantly increased expression of SLC22A17, which is the NGAL receptor, in neutrophils from patients with IAs. This may reflect a possible interaction with circulating NGAL. However, we did not observe significantly higher levels of NGAL or MPO in circulating neutrophils, suggesting that these proteins may originate from the aneurysm sac itself, or other circulating cells.

[0112] Further analysis of the expression data described in this Example supports an association between activated circulating neutrophils and IA presence. Gene set enrichment analysis reveals that neutrophils from IA patients have higher levels of gene expression associated with leukocyte activation. This is evidenced by increased expression levels of several CD antigens from the "leukocyte activation" ontology (CD ID, CD7, CD86, and CD247) as well as CD 177, a marker of neutrophil activation. IPA also reveals functions indicative of activated neutrophils, showing networks consistent with activation of cellular movement, cell-to-cell signaling, and cell proliferation. The fact that neutrophil expression data segregated aneurysms by size in PCA and MDS (Fig 3 A and B) may indicate a correlation between the degree of IA advancement and neutrophil activation. The present disclosure thus indicates that peripheral neutrophil activation may play a role in IA development. Specifically, in this Example, we identified an IA-associated RNA expression signature of 82 transcripts in circulating neutrophils. This signature demonstrated a statistical power >0.80 and was able to distinguish patients with IAs from paired controls in several analyses. These transcripts also separated patients with IAs from unpaired controls in a small population. [0113] Example 1, S I Table. Primers used for qPCR and their efficiencies.*

*Primers were selected using Primer3 and NCBFs Primer Blast. All efficiencies were within the range of 0.90- 1.10. (bp=base pair, Eff=efficiency, Prod.=product, Temp.=temperature) [0114] Example 1, S2 Table. RNA Quality.*

ID Class 260/280 RIN

Discovery Cohort

CI Control 2.07 7.0

C2 Control 2.02 7.4

C3 Control 2.05 6.6

C4 Control 1.92 6.5

C5 Control 1.99 7.1

C6 Control 2.08 7.1

C7 Control 2.05 7.9

C8 Control 1.96 6.7

C9 Control 2.08 6.2

CIO Control 2.04 7.3

Cl l Control 1.97 6.4

Al Aneurysm 2.04 7.8

A2 Aneurysm 2.07 7.5

A3 Aneurysm 2.06 8.1

A4 Aneurysm 2.02 6.1

A5 Aneurysm 2.03 7.3

A6 Aneurysm 1.99 6.5

A7 Aneurysm 2.02 7.5

A8 Aneurysm 2.05 7.7

A9 Aneurysm 2.07 6.0

A10 Aneurysm 1.95 7.2

Al l Aneurysm 1.97 6.9

Replication Cohort

C12 Control 2.03 6.0

C13 Control 2.04 6.6

C14 Control 2.05 6.0

C15 Control 2.02 8.2

C16 Control 2.04 6.0

A12 Aneurysm 2.06 6.0

A13 Aneurysm 1.97 6.4

A14 Aneurysm 2.00 6.9

A15 Aneurysm 1.96 7.4

A16 Aneurysm 2.03 7.2

* The quality of the RNA samples was assessed by the 260/280 ratio and the RIN. (RIN=RNA integrity number)

[0115] Example 1, S3 Table. Characteristics of 16 intracranial aneurysms in the group of 11 patients with IAs (3 patients had multiple intracranial aneurysms)* Example 1, S4 Table. RNA Sequencing Quality Control Analysis.* IA Family

Size IA Presence of History of

ID (mm) Location Additional IAs IA Indications for DSA

MRI for hand numbness indicated

Al 10 VB junction No No possible IA

A2 8 Ophthalmic No No Follow-up imaging of known IA

Incidental finding on CT

A3 4.5 MCA No No indicated possible IA

Yes: +2 (1.5 mm and MRI for headache indicated

A4 4 Ophthalmic 3 mm ICA) No possible IA

Yes: +2 (2.3 mm Incidental finding on MRI

A5 10.8 MCA MCA, small AComA) No indicated possible IA

A6 9 PComA No No Follow-up of known IA

MRA and CT for tremor revealed

A7 5 BT No No possible IA

MRI for decreased vision in left

A8 13 ACA No Yes eye indicated possible IA

MRI for double vision indicated

A9 19 ICA No No possible IA

Yes: +1 (3.5 mm MRI for tremors indicated

A10 5 ICA ICA) Yes possible IA

MRI for headache indicated

Al l 3 BT No No possible IA

* Aneurysm size ranged from 1.5mm to 19mm. Seven of 16 IAs (44%) were classified as small (greatest diameter <5mm) and 9 (56%) were classified as large (greatest diameter >5 mm). The aneurysms were situated at various locations in the Circle of Willis, with most being around the internal carotid artery (ICA) and its branches. Two patients with IAs had a family history of the disease. In general, digital subtraction angiography was performed for either confirmation of IA presence after an incidental finding of IA on noninvasive imaging, or for follow-up imaging of a previously detected IA. (ACA=anterior cerebral artery, AComA=anterior communicating artery, BT=basilar terminus, CT=computed tomography, DSA=digital subtraction angiography, IA=intracranial aneurysm, ICA=internal carotid artery, MCA=middle cerebral artery, MRA=magnetic resonance angiography, MRI=magnetic resonance imaging, PComA=posterior communicating artery,

VB=vertebrobasilar)

[0116] Example 1, S4 Table. RNA Sequencing Quality Control Analysis. *

Poor

M. Qual. Seqs. % % M. Detected

ID Class Seqs. Seqs. Length GC Aligned Aligned Transcripts discovery Cohort

CI Control 27 0 51 48 96.50 24.8 17050

C2 Control 59.3 0 51 49 96.50 54.4 17930

C3 Control 35.9 0 51 49 94.90 32.3 18291

C4 Control 68.9 0 51 49 96.60 63.3 18465

C5 Control 53.3 0 51 50 95.80 48.5 17961

C6 Control 80.3 0 51 50 96.30 73.5 18418

C7 Control 97.4 0 51 50 96.00 88.9 18066

C8 Control 67.8 0 51 49 96.40 62.2 18714

C9 Control 36.2 0 51 49 96.10 33.2 17966

CIO Control 79.3 0 51 49 96.80 72.7 17366 Cl l Control 89.1 0 51 50 95.70 81.1 18479

Al Aneurysm 48.1 0 51 49 96.90 44.5 18003

A2 Aneurysm 35.7 0 51 48 96.80 32.9 17223

A3 Aneurysm 60.4 0 51 49 96.90 55.9 18662

A4 Aneurysm 55.9 0 51 49 96.90 51.4 18155

A5 Aneurysm 61.3 0 51 49 97.20 56.9 17682

A6 Aneurysm 23.4 0 51 49 95.90 21.4 17366

A7 Aneurysm 35.8 0 51 49 96.40 32.3 18625

A8 Aneurysm 26.9 0 51 50 97.10 24.9 16401

A9 Aneurysm 29.1 0 51 50 96.60 26.8 17311

A10 Aneurysm 59.5 0 51 49 96.00 54.1 18617

Al l Aneurysm 14.6 0 51 49 95.00 13 17401

Replication Cohort

Cll Control 74.4 0 51 50 95.10 66.8 18329

C13 Control 66 0 51 50 95.50 60 18579

C14 Control 58.5 0 51 51 95.50 53 18273

C15 Control 64.3 0 51 51 94.90 57.9 18448

C16 Control 47.3 0 51 50 95.30 42.7 18039

A12 Aneurysm 42.2 0 51 51 95.50 38.3 17697

A13 Aneurysm 75.3 0 51 51 95.90 68.6 18313

A14 Aneurysm 64.9 0 51 50 96.60 59.5 17546

A15 Aneurysm 39.8 0 51 50 97.40 36.8 18031

A16 Aneurysm 42.1 0 51 50 95.20 38.2 17330

* The quality of the RNA sequencing experiments was measured. Overall, prior to alignment all samples had an average of 53.75 M sequences. MultiQC reported that the sequencing experiments had an average of 49.09 M mapped reads with a 96.13% read mapping rate, and detected an average of 17259 transcripts (transcripts with FPKM>0). (Align. =alignment, M.=million, Seqs.=sequences, Qual.=quality)

[0117] Example 1, S5 Table. Transcripts involved in the 4 networks constructed by

Ingenuity Pathway Analysis (IP A). *

Focus Top Diseases and

Network Molecules in Network P-Score Molecules Functions

AKR1C3, Apl, BCR (complex), 41 Carbohydrate CCL23, CD163, CXCL8, Collagen(s), Metabolism, DGKH, ERK1/2, FBN1, FCRL5, Cardiovascular Fibrinogen, Gm-csf, ITGA7, IgG, System IgGl, Immunoglobulin, Integrin, Development and KIR2DS4 (includes others), KLRC2, Function, Cellular Laminin, Mek, NOG, Nrlh, RCVRN, Movement

SERPINF2, SERPING1, SLC12A7,

TCL1A, THBS1, Tgf beta, VLDLR,

VWF, elastase, trypsin

B ADTRP, APP, B4GALNT3, CACHE ⁾ 1, 30 15 Cell-To-Cell

CDHR2, ECI2, ETHE1, ETV7, Signaling and FAM90A1, GBP5, G PTAB, HPCA, Interaction, Drug HTR2C, IL10RA, KIAA1598, Metabolism, KLHL41, MAOA, MAOB, MY07B, Energy Production MYPN, NEB, PPARA, PRSS21,

PRUNE2, Ppap2a, RPL39L, RPL7L1, SCAMP5, SDC3, SRC, STYXL1,

TMOD4, UBC, VWA8, XIRP2

ADAMTS1, ADCY, 30 15 Lipid Metabolism, AKR1 C 1/AKRl C2, AMPK, Akt, Small Molecule CYP4F2, Cg, EGR2, ERK, FADS2, Biochemistry, FBXW8, FLT3, FSH, Focal adhesion Behavior kinase, G0S2, GNLY, Hdac, Hsp70,

Hsp90, IL1, Insulin, KCNMA1, Lh,

NRG1, PDGF BB, PID1, PTGDS, Pka,

Pkc(s), Ras, SCT, SLC22A17, Vegf,

caspase, p85 (pik3r)

D AK5, AKR1 C 1/AKRl C2, ARFGAP3, 23 Organismal Injury

ASS1, BATF2, C1R, C4BPA, CALB2, and Abnormalities, CAPG, CARD17, CCL23, CD177, Tissue CD276, CEBPA, CSF3, CTN B 1, Morphology, CTN D2, Cebp, Collagen type VI, Cellular Growth DACT1, DACT2, DACT3, DSPP, and Proliferation GPC4, HLA-F, HRK, IFNG, KLK8,

LILRA2, OSM, PDCD1LG2, PRTN3,

SLC16A1, VLDLR, WNT5A

A table of the names of transcripts included in the top 4 networks derived from IP A, as well as the top diseases and functions of these transcripts. Neutrophil transcripts in bold were of the 82 differentially expressed

transcripts between patients with and without IA (p-value<0.05). Each network's p-score was derived from its p- value [p-score = -LoglO (p-value)] calculated by the Fisher's exact test. For a network with a p-score of 10, the odds of generating this network by chance alone is less than 1 out of 10 ¹⁰ .

[0118] Example 1, S6 Table. Clinical characteristics of the unpaired cohort of 5

patients with intracranial aneurysms and 5 control subjects without intracranial aneurysms

(confirmed on imaging)*

Patients with IA Patients without

(n=5) IA (n=5)

Age (years) (Mean ± SE) 56.8±3.95 48.8±6.65

Age (years) [Median (Ql/Q3)l 56 (53/57) 54 (51/55)

Sex

Female 60% 20%

Current smoker

Yes 40% 0%

Comorbidities

Hypertension 40% 40%

Hyper lipidemia 20% 20%

Heart disease 20% 0%

Previous stroke 0% 0% Diabetes mellitus 20%

Osteoarthritis 20%

(IA=intracranial aneurysm, SE=standard error, Q=quartile)

[0119] Example 1, S7 Table. Characteristics of 6 intracranial aneurysms in the replication group of 5 patients with IAs (one patients had multiple intracranial aneurysms)*

* Aneurysm size ranged from 3.5 mm to 7 mm. Five of 6 IAs (83%) were classified as small (greatest diameter <5mm) and 1 (17%) was classified as large (greatest diameter >5 mm). The aneurysms were situated at various locations in the Circle of Willis, with most being in the anterior vasculature (ACA and MCA). (ACA=anterior cerebral artery, AComA=anterior communicating artery, BT=basilar terminus, CT=computed tomography, DSA=digital subtraction angiography, IA, intracranial aneurysm, MCA=middle cerebral artery, MRA=magnetic resonance angiography, MRI=magnetic resonance imaging)

Example 2

[0120] This Example extends the disclosure of Example 1. In particular, in this Example, neutrophil RNA extracted from blood samples from 40 patients (20 with unruptured IA, 20 IA-free controls) was subjected to next-generation RNA sequencing to obtain neutrophil transcriptomes. In a randomly selected training cohort of 30 samples (n=15 with IA and n=15 controls), we performed differential expression analysis. Significantly differentially expressed transcripts (False Discovery Rate (FDR)<0.05, fold change of >1.5) were used to construct prediction models for IA using four known supervised machine learning approaches (linear discriminant analysis, k-nearest neighbors, nearest shrunken centroids, and support vector machines). These models were tested in the remaining 10 patients (testing cohort) and their performance was assessed by receiver-operating- characteristic curves. Real-time PCR was used to corroborate expression differences of 7 model transcripts in neutrophil samples from a new, independent cohort (n=10). The training cohort yielded 26 highly significantly differentially expressed neutrophil transcripts. Models using these transcripts identified IA patients in the testing cohort with accuracy ranging from 0.60 to 0.90. The best performing model was a diagonal linear discriminant analysis classifier (Area Under the Curve (AUC)=0.80 and accuracy=0.90). Six of seven differentially expressed genes were confirmed by quantitative PCR using isolated neutrophils from a separate patient cohort. Thus, this Example demonstrates use of machine learning methods to classify IA cases and create predictive models for unruptured IAs using circulating neutrophil transcriptome data.

[0121] The following materials and methods were used to produce the results described in this Example.

[0122] Cohort Generation

[0123] Study Population

[0124] This study was approved by the University at Buffalo Health Sciences

Institutional Review Board. Methods were carried out in accordance with the approved protocol. Written informed consent was obtained from all subjects. 106 peripheral blood samples were collected from patients undergoing cerebral digital subtraction angiography (DSA) at Gates Vascular Institute in Buffalo, NY: 51 patients had a positive IA diagnosis and 55 had a negative IA diagnosis (controls). Positive or negative IA diagnosis was confirmed by imaging, and patient medical records were collected.

[0125] Patients undergoing cerebral digital subtraction angiography (DSA) with positive and negative intracranial aneurysm (IA) diagnoses were enrolled in this study.

Reasons for the patients to receive DSA included confirmation of findings from noninvasive imaging of the presence of IAs, vascular malformations, or carotid stenosis, or follow-up noninvasive imaging of previously detected IAs. All consenting patients were older than 18 years, were English speaking, and had not received previous treatment for IA. We excluded patients who potentially had altered immune systems; this included patients who were pregnant, had recently undergone invasive surgery, were undergoing chemotherapy, had a body temperature above 37.78°C (100°F), had received solid organ transplants, had autoimmune diseases, and those who were taking prednisone or any other immunomodulating drugs. Furthermore, included patients did not have any other known cerebrovascular malformations or extracranial aneurysms, including abdominal aortic aneurysms.

[0126] Neutrophil isolation

[0127] Sixteen mL of blood was drawn from the access catheter in the femoral artery and transferred into two 8 mL, citrated, cell preparation tubes (BD, Franklin Lakes, NJ). Neutrophils were isolated within 1 hour of peripheral blood collection, as described elsewhere (Jiang K, Sun X, Chen Y, Shen Y, Jarvis JN (2015) RNA sequencing from human neutrophils reveals distinct transcriptional differences associated with chronic inflammatory states. BMC Med Genomics 8: 55). Cell preparation tubes were centrifuged at 1,700 ^χ g for 25 minutes to separate erythrocytes and neutrophils from mononuclear cells and plasma in the peripheral blood samples via a Ficoll density gradient. Erythrocytes and neutrophils were collected into a 3 mL syringe. Following hypotonic lysis of red blood cells, neutrophils were isolated by centrifugation at 400 x g for 10 min and disrupted and stored in TRIzol reagent (Life Technologies, Carlsbad, CA) at -80°C until further processing. Neutrophils isolated in this fashion are more than 98% CD66b+ by flow cytometry and contain no contaminating CD14+ monocytes. (Jarvis JN, Dozmorov I, Jiang K, Frank MB, Szodoray P, et al. (2004) Novel approaches to gene expression analysis of active polyarticular juvenile rheumatoid arthritis. Arthritis Res Ther 6: R15-r32.)

[0128] RNA preparation

[0129] Neutrophil RNA was extracted using TRIzol, according to the manufacturer's instructions. Trace DNA was removed by DNase I (Life Technologies, Carlsbad, CA) treatment. RNA was purified using the RNeasy MinElute Cleanup Kit (Qiagen, Venlo, Limburg, Netherlands) and suspended in RNase-free water. The purity and concentration of RNA in each sample was measured by absorbance at 260 nm on a NanoDrop 2000 (Thermo Scientific, Waltham, MA), and 200-400 ng of RNA was sent to our university's Next- Generation Sequencing and Expression Analysis Core facility for further quality control. Precise RNA concentration was measured at the core facility via the Quant-iT RiboGreen Assay (Invitrogen, Carlsbad, CA) with a TBS-380 Fluorometer (Promega, Madison, WI), whereas the quality of the RNA samples was measured with an Agilent 2100 BioAnalyzer RNA 6000 Pico Chip (Agilent, Las Vegas, NV). RNA samples of acceptable purity (260/280 ratio of >1.9) and integrity (RIN > 5.0) were considered for RNA sequencing.

[0130] RNA sequencing

[0131] RNA libraries were constructed using the Illuminal TruSeq RNA Library

Preparation Kit (Illumina, San Diego, CA). All samples were subjected to 50-cycle, single- read sequencing in the HiSeq2500 (Illumina) and were demultiplexed using Bcl2Fastq v2.17.1.14 (Illumina). Gene expression analysis was carried out using the Tuxedo Suite. For each sample, short RNA fragment data in the FASTQ format was compiled and aligned to the human reference genome (human genome 19 - hgl9) using TopHat v2.1.13. To evaluate the quality of RNA sequencing, we performed quality control analysis using FASTQC and visualized and compared the aggregate quality control data using MultiQC.

[0132] Transcript expression levels were calculated from counts using Transcripts Per

Million (TPM) normalization for comparison of RNA levels between samples. Samples were processed in two batches. Therefore, we performed batch effect correction using ComBat under the default settings in R. This was performed on expression data for all transcripts with an average TPM .O in at least one of the two groups. (See Example 2, Supplemental Table 1 for batch information).

[0133] Differential expression analysis

[0134] Prior to differential expression analysis, neutrophil transcriptomes were randomly divided into two cohorts: a training cohort (n=30) and a testing cohort (n=10), each containing half IA and half control transcriptomes. Differential gene expression analysis in the training cohort was carried out using F statistics to assess differential variation about the mean on a transcript-by-transcript basis. Multiple testing correction was performed by using the Benjamini-Hochberg method, and q-values were reported for each transcript. Transcripts were considered significantly differentially expressed at an FDR-adjusted p-value (q- value)<0.05.

[0135] Bioinformatics

[0136] We performed gene ontology term enrichment analysis using the open source

Gene Ontology enRIchment anaLysis and visuaLizAtion tool (GORILLA) on all

differentially expressed transcripts (q<0.05). This was done using a background gene list of neutrophil RNA expression (average fragments per kilo base of transcript per million

(FPKM) mapped reads, FPKM .0) of 3 healthy individuals, described elsewhere. (Jiang K, Sun X, Chen Y, Shen Y, Jarvis JN (2015) RNA sequencing from human neutrophils reveals distinct transcriptional differences associated with chronic inflammatory states. BMC Med Genomics 8: 55.) This tool identified gene ontology (GO) terms that are enriched in genes with increased or decreased expression in IA compared to the background neutrophil expression using standard hyper geometric statistics. Associated gene ontology processes and functions were reported if the enrichment FDR adjusted p-value (q-value)<0.20 (20% FDR).

[0137] Classification Model Development

[0138] Feature selection

[0139] Prior to model training, the set of differentially expressed transcripts was reduced by filtering. We retaining only transcripts with an FDR<0.05 and absolute fold- change >1.5. To visualize how the remaining transcripts separated IA from control samples, we performed principal component analysis (PCA) in R using the prcomp package under the default settings.

[0140] Model training [0141] Using the selected transcripts, we trained classification models using

MATLAB Statistics and Machine Learning Toolbox and R bioconductor. Specifically, we considered four algorithms that have been successfully used for disease classification from gene expression data. These methods included K-Nearest Neighbors Classification, (Baker CJ, Fiore A, Connolly ES, Jr., Baker KZ, Solomon RA (1995) Serum elastase and alpha- 1- antitrypsin levels in patients with ruptured and unruptured cerebral aneurysms. Neurosurgery 37: 56-61; discussion 61-5) Linear Discriminant Analysis, (Phillips J, Roberts G, Bolger C, el Baghdady A, Bouchier-Hayes D, et al. (1997) Lipoprotein (a): a potential biological marker for unruptured intracranial aneurysms. Neurosurgery 40: 1112-1115; discussion 1115-1117), Nearest Centroids Classification, (Sandal cioglu IE, Wende D, Eggert A, Regel JP, Stolke D, et al. (2006) VEGF plasma levels in non-ruptured intracranial aneurysms. Neurosurg Rev 29: 26-29), and Support Vector Machines (SVM). (Zhang HF, Zhao MG, Liang GB, Song ZQ, Li ZQ (2013) Expression of pro-inflammatory cytokines and the risk of intracranial aneurysm. Inflammation 36: 1195-1200). Each method was applied to the training cohort separately and evaluated with a leave-one-out (LOO) cross-validation to estimate model performance and prevent overfitting. Nearest neighbors classification: The k-nearest neighbors method (Chalouhi N, Theofanis T, Starke RM, Zanaty M, Jabbour P, et al. (2015) Potential role of granulocyte-monocyte colony-stimulating factor in the progression of intracranial aneurysms. DNA Cell Biol 34: 78-8) with a cosine metric (Cosine Nearest Neighbors (cosine NN)) was employed. The number of neighbors, k, was set as 5 for cosine NN. The resulting model classified test samples by calculating their distance to each training sample. The test samples' labels were predicted by majority vote, choosing the class that was most common among their k nearest neighbors. Linear discriminant analysis: We trained a classifier using diagonal linear discriminant analysis (DLDA) as described elsewhere. (Baker CJ, Fiore A, Connolly ES, Jr., Baker KZ, Solomon RA (1995) Serum elastase and alpha- 1 -antitrypsin levels in patients with ruptured and unruptured cerebral aneurysms. Neurosurgery 37: 56-61; discussion 61-52). This method seeks the linear combination of transcripts which best separate two classes using a diagonal covariance matrix. The linear model coefficients associated with transcripts (discriminant scores) relayed importance of each transcript to the prediction model. (Chen J, Han L, Xu X, Tang H, Wang H, et al. (2015) Serum biomarkers VEGF-C and IL-6 are associated with severe human Peripheral Artery Stenosis. J Inflamm (Lond) 12: 50). Classification was performed by projecting a test sample onto the maximally separating direction which was determined by discriminant scores and calculating the corresponding posterior probability of IA. Nearest centroids classification: We used a modification of the nearest centroids technique, called Nearest Shrunken Centroids (NSC). (Sandalcioglu IE, Wende D, Eggert A, Regel JP, Stolke D, et al. (2006) VEGF plasma levels in non-ruptured intracranial aneurysms. Neurosurg Rev 29: 26-29). This method calculates class-specific centroids (standard deviation normalized averages) for each transcript and refined them by eliminating those with variable expression. Classification was performed by comparing the expression of the included model transcripts with the centroids of the two classes and assigning it to a class that is closest to in squared distance. (Sandalcioglu IE, Wende D, Eggert A, Regel JP, Stolke D, et al. (2006) VEGF plasma levels in non-ruptured intracranial aneurysms. Neurosurg Rev 29: 26-29). Support vector machine: The most complex classification algorithm we implemented was SVM. (Kimura H, Okada O, Tanabe N, Tanaka Y, Terai M, et al. (2001) Plasma monocyte chemoattractant protein- 1 and pulmonary vascular resistance in chronic thromboembolic pulmonary hypertension. Am J Respir Crit Care Med 164: 319-324). To separate the binary labeled training samples, SVM finds a hyper-plane which is maximally distant from samples of either class. A linear kernel was used in model creation. The resulting model classified test samples by mapping them to a higher-dimensional space and making decisions based on their signed distance to the hyper- plane.

[0142] Classification Model Evaluation

[0143] Model assessment in the training cohort

[0144] The performance of each model in the training cohort was estimated using the results of the LOO cross-validation. The model classifications were compared to each patient's clinical diagnosis from imaging, and the true positives (TP), true negatives (TN), false positives (FP), and false negatives (FN) were counted. Each model's performance was first assessed by calculating the model's sensitivity, specificity, and accuracy:

Sensitivity = TP /(TP + FN)

TN

Specificity =————

' ^y TN + FP

Accuracy = (TP + TN)/(TP + FP + FN + TN)

[0145] Based on model predictions, we created receiver operating characteristic

(ROC) curves and calculated the area under the ROC curve (AUC) to assess model performance.

[0146] Validation of the models in an independent testing cohort

[0147] Classification models were independently tested on transcriptomes from the testing cohort. TPM values of these model features were input into the models for classification of IA presence. The classification results were compared to clinical diagnoses to calculate the true sensitivity, specificity, and accuracy for each model. ROC curves were constructed and AUCs were used to assess the performance of each classifier.

[0148] Cross-validation over all samples

[0149] Since the models were fit using data points from a randomly selected training dataset (n=30), selection bias may introduce inconsistency in model predictions. To increase the models' reliability of prediction and to create algorithms more generalizable to the entire population, we implemented a LOO cross-validation using the expression levels of the 26 selected transcripts from all 40 patients for each model. This essentially retrained the models in 40 different training sets consisting of 39 samples and performed testing on the remaining sample. As before, classification results were used to calculate the sensitivity, specificity, and accuracy for each model, as well as find the AUC of the ROC curve for each modified classifier.

[0150] Positive and negative predictive values of the models

[0151] We further assessed the predictive value of the classification models by calculating their positive predictive values (PPV) and negative predictive values (NPV). (Li P, Zhang Q, Wu X, Yang X, Zhang Y, et al. (2014) Circulating microRNAs serve as novel biological markers for intracranial aneurysms. J Am Heart Assoc 3 : e000972), PPVs and PVs were estimated using the following formulas based on the Bayes' theorem:

Sensitivity x Prevalence

Sensitivity x Prevalence + (1— Specif icity)x (1— Prevalence)

Specificity x (1— Prevalence)

NPV =

(1— Sensitivity)x Prevalence + Specificity x (1— Prevalence)

[0152] The PPV and NPV were calculated over a range of prevalence from 0-100%, noting the reported range of IA prevalence (3.2%-7%) according to established approaches: (Meng H, Tutino VM, Xiang J, Siddiqui A (2014) High WSS or low WSS? Complex interactions of hemodynamics with intracranial aneurysm initiation, growth, and rupture: toward a unifying hypothesis. ATNR Am J Neuroradiol 35: 1254-1262; Chalouhi N, Points L, Pierce GL, Ballas Z, Jabbour P, et al. (2013) Localized increase of chemokines in the lumen of human cerebral aneurysms. Stroke 44: 2594-2597; Frosen J, Tulamo R, Paetau A,

Laaksamo E, Korja M, et al. (2012) Saccular intracranial aneurysm: pathology and mechanisms. Acta Neuropathol 123 : 773-786; Yu L, Wang J, Wang S, Zhang D, Zhao Y, et al. (2017) DNA Methyl ati on Regulates Gene Expression in Intracranial Aneurysms. World Neurosurg 105: 28-36).

[0153] Validation of expression differences by qPCR in an independent cohort

[0154] To validate expression differences in the 26 model genes, quantitative polymerase chain reaction (qPCR) was performed. Due to limitations in mRNA volume, qPCR was performed on 7 model transcripts in 10 additional patients (5 with IA and 5 controls), as described in Example 1. In brief, oligonucleotide primers were designed with a 60°C melting temperature and a length of 15-25 nucleotides to produce PCR products with lengths of 50-250 base pairs using Primer3 software and Primer BLAST (NCBI, Bethesda, MD). The replication efficiency of each primer set was tested by performing qPCR on serial dilutions of cDNA samples (primer sequences, annealing temperatures, efficiencies, and product lengths are shown in Example 2, Supplemental Table 2).

[0155] For reverse transcription, first-strand cDNA was generated from total RNA using Omni Script Reverse Transcriptase kit (Qiagen, Venlo, Limburg, Netherlands) according to the manufacturer's directions. Quantitative PCR was run with 10 ng of cDNA in 25 μΐ. reactions in triplicate in the Bio-Rad CFX Connect (Bio-Rad, Hercules, California) using ABI SYBR Green I Master Mix (Applied Biosystems, Foster City, California) and gene-specific primers at a concentration of 0.02 μΜ each. The temperature profile consisted of an initial step of 95°C for 10 min, followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 min, and then a final melting curve analysis from 60°C to 95°C over 20 min.

[0156] Gene-specific amplification was demonstrated by a single peak using the Bio-

Rad dissociation melt curve. Samples were normalized based on GAPDH, 18s rRNA, and GPI expression, which was run in parallel reactions to the genes of interest. These values were used to calculate average fold-change between the two groups using the 2 ^"ΔΔα method. (Leopold JA (2015) The Central Role of Neutrophil Gelatinase-Associated Lipocalin in Cardiovascular Fibrosis. Hypertension 66: 20-22). These values were calculated for each housekeeping gene and averaged. Average fold-change in gene expression measured by qPCR data in the new cohort was then compared to the fold-change calculated from RNA sequencing in the training cohort.

[0157] The following results were obtained using the foregoing materials and methods described in this Example.

[0158] Study participants

[0159] During the study period, we collected 106 blood samples (51 from patients with IA, 55 from control subjects) as well as angiographic images and medical records data from individuals undergoing cerebral digital subtraction angiography (DSA). Of the blood samples collected, 43 (20 from IA patients, 23 from controls) met our criteria and also had neutrophil RNA of sufficient quality and volume for sequencing. A total of forty patients (20 with IA and 20 controls) were then chosen and randomly divided into a 30 patient training cohort (n=15 IA and n=15 control) and a 10 patient testing cohort (n=5 IA and n=5 control). See Example 2, Table 1 for the characteristics of the two cohorts. These samples were of sufficient quality and had an average 260/280 of 2.02 (range 1.90-2.12) and an average RNA integrity number (RIN) of 6.88 (range 5.2-8.2) (Example 2, Supplemental Table 3). Patients with IAs had aneurysms ranging in size from 1-19 mm and included 5 individuals with multiple IAs (see Example 2, Supplemental Table 4). A portion of these samples (n=22) were described in Example 1, which analyzed neutrophil expression differences between patients with and without IA.

[0160] Differential RNA expression in neutrophils from patients with IA vs. control

[0161] RNA sequencing data was used to identify differentially expressed neutrophil transcripts between the 15 patients with IA and 15 controls in the training cohort. Overall, our sequencing experiments had an average of 53.84 million sequences per sample and a 95.4% read mapping rate (% aligned) (see Example 2, Supplemental Table 5). The volcano plot in Figure 7A shows neutrophil expression differences between IA patients and controls in terms of average fold-change in expression and significance level. From 12,775 transcripts with average TPM .O in either group, differential expression analysis identified 95 transcripts that were significantly differentially expressed (q<0.05). Gene set enrichment analysis performed using these 95 differentially expressed transcripts showed genes with higher levels in the IA group were involved in defense response, leukocyte activation, stem cell maintenance, maintenance of cell number, cell activation, and stem cell development. Genes with lower levels in IAs were involved in regulation of glutathione and tetrapyrrole binding (Example 2, Table 2).

[0162] Selected transcripts for model training

[0163] Prior to model training, we performed feature selection by filtering to identify disease-related transcripts and reduce the data dimensionality to facilitate downstream analysis. Our statistical criteria of false discovery rate (FDR)<0.05 and an absolute fold- change > 1.5 resulted in retention of the 26 transcripts that are demonstrated in shading in Figure 7A and listed in Example 2, Table 3. The PCA in Figure 7B shows that these 26 transcripts could generally discriminate patients with IAs from the controls. The top three principal components represented 47.8% of the variation; PCI contained 22.4% variation, PC2 contained 15.3% variation, and PC3 contained 10.1% variation. Overall, 60% of the aneurysm samples and 80% of the control samples could be grouped together by PCA.

[0164] Classification models of IA have high performance in training and testing datasets

[0165] Using the expression of these 26 transcripts, we trained four classification models, using cosine nearest neighbors (cosine NN), diagonal linear discriminant analysis (DLDA), nearest shrunken centroids (NSC), and support vector machines (SVM). Figure 7C shows the sensitivity, specificity, and accuracy of the models, which were estimated from LOO cross-validation. There was moderate performance by each classification method, with accuracies that ranged from 0.50 to 0.73. Evaluation by ROC curve analysis showed a range in AUCs from 0.54 to 0.72 (Figure 7D) across all methods. In this dataset, DLDA performed the best, with a sensitivity of 0.67, a specificity of 0.80, an accuracy of 0.73, and an AUC of 0.72.

[0166] To independently validate the models, we implemented them in the testing cohort neutrophil transcriptomes from 10 patients. The PCA in Figure 8 A shows that the 26 transcripts could discriminate patients with IAs from controls in the testing cohort as well.

Overall, 100% of the aneurysm samples and 80% of the control samples could be grouped together by PCA. In the testing cohort, the models predicted the aneurysm status with a range in accuracy from 0.60 to 0.90 (Figure 8B). The ROC analysis in Figure 8C demonstrates that the models' AUCs ranged from 0.62 to 0.80. In this cohort, the DLDA classification model again performed the best, with a sensitivity of 0.80, specificity of 1.0, an accuracy of 0.90, and an AUC of 0.80.

[0167] Cross-validation to increase model reliability

[0168] To increase the models' reliability, we employed LOO cross-validation using all patient transcriptomes and refit them in all 40 datasets (all combinations of 39 training samples and 1 testing sample). This analysis revealed the models' accuracy ranged from 0.63 to 0.80 (Figure 9A) and their AUCs ranged from 0.69 to 0.84 (Figure 9B). Again, the DLDA model performed the best, with a sensitivity of 0.65, specificity of 0.95, accuracy of 0.80, and an AUC of 0.84.

[0169] Models have high negative predictive value

[0170] Given their range of performance, we wanted to know how useful the models would be at detecting IA. Their value, however, would be inherently influenced by the prevalence of IA in a given target population. To estimate this, we plotted the positive predictive value (PPV) and negative predictive value (NPV) for each model (Figure 9C and D) using the sensitivity and specificity reported after the LOO cross-validation in all datasets. The rate aneurysm incidence found in the published literature ranged from 3.2% (Meng H, Tutino VM, Xiang J, Siddiqui A (2014) High WSS or low WSS? Complex interactions of hemodynamics with intracranial aneurysm initiation, growth, and rupture: toward a unifying hypothesis. AJNR Am J Neuroradiol 35: 1254-1262) to 7%, (Chalouhi N, Points L, Pierce GL, Ballas Z, Jabbour P, et al. (2013) Localized increase of chemokines in the lumen of human cerebral aneurysms. Stroke 44: 2594-2597) and is indicated as shown on the graph. At a 5% incidence of aneurysm, the models' PPV ranged from 0.10 to 0.41 and PV ranged from 0.96 to 0.98. The DLDA classifier had the highest PPV (0.41) and NPV (0.98).

[0171] Independent validation of expression differences by RT-qPCR

[0172] We performed a corroboration study to determine if the differential expression of 7 model genes could be detected in a new population of IA and control patients. We used samples from 10 additional patients (5 with IA and 5 controls) from which we collected neutrophil RNA but did not sequence (see Example 2, Supplemental Table 6 for patient information for this cohort and Example 2, Supplemental Table 7 for aneurysm information from the IA patients in this cohort). These samples were used for quantitative polymerase chain reaction (qPCR) analysis differentially of CD177, CYP1B1, ARMC12, OLAH, CD163, G0S2, and FCRL5, which were selected because they were also expressed as described in Example 1. Figure 10 shows the qPCR results of this corroborative study in comparison with expression differences obtained from RNA sequencing in the training cohort. Six of the 7 genes demonstrated average fold-change in the same direction and of similar magnitude to those in the original cohort. This indicated that the difference in expression of these transcripts between patients with IA and control is consistent.

[0173] It will be recognized from the foregoing description in this Example that Neutrophils play a role in the progressive inflammation that typifies IAs. Therefore, we analyzed whether gene expression patterns in circulating neutrophils could reflect the presence of the disease. As described in Example 1, we found that patients with unruptured IAs and aneurysm-free controls had significant RNA expression differences in circulating neutrophils. In this Example, we analyzed biomarkers using neutrophil RNA expression levels from blood samples to predict IA presence. We used RNA expression profiling in circulating neutrophils to identify 26 transcripts that were highly associated with the presence of IA. Machine learning algorithms were then implemented to develop classification models that used these 26 transcripts to predict the presence of aneurysm.

[0174] Classification models of IA [0175] Without intending to be constrained by any particular theory, it is believed this disclosure is the first to demonstrate IA prediction from RNA expression patterns in the blood. The four models we trained performed with an average classification accuracy of 80%, a level which indicates promise for further investigation of RNA expression biomarkers for IA. Overall, classification by DLDA achieved the best performance in our data. This model, consistently had the highest accuracy and AUC over multiple analyses, including cross- validation during model training (accuracy=0.73, AUC=0.72), independent model validation (accuracy=0.90, AUC=0.80) and cross-validation across the entire dataset (accuracy=0.80, AUC=0.84). See Example 2, Supplemental Table 8 for a detailed summary of the

performance of all models.

[0176] The DLDA classifier may have performed better than other methods because it accommodated inter-sample variability in the 26 transcripts. While average expression of each transcript is significantly different between IA and controls, their expression levels may vary between individuals, and thus modeling techniques that broadly survey patterns of gene expression may afford better IA classification. Without intending to be constrained by any particular theory, it is considered that the DLDA method may have functioned as described herein because: (1) it ranks transcripts by importance, giving more weight to the most consistently informative transcripts (unlike non-parametric approaches such as nearest neighbors). And (2) it uses information from all transcripts to project test samples to the direction which best separate the classes. Thus, a linear combination of transcripts can generate a stable IA prediction, and accommodate for inter-sample variability. Additionally, ignoring correlations between genes as DLDA does, can provide a simple model and produced lower misclassification rates than more sophisticated classifiers, such as SVM.

[0177] In this disclosure, classifiers were developed based on 30 transcriptomes that were randomly selected from all available data (n=40). Randomization was used so we could test the viability of IA prediction in patients that have potentially confounding covariates (comorbidities and demographics). Example 2, Table 1 shows that, in the training cohort, there were differences in the rate of smoking between the IA and control groups, which may reflect an established association between IA and smoking. The approach in this Example was designed differently than as described in Example 1, in which we found an 82-transcript expression signature of IA by transcriptome profiling of a cohort-controlled group of patients. But even with this difference, 38% (10) of 26 classifier transcripts were part of the 82-transcript signature described herein. These include CYP1B1, CD177, ARMC12, OLAH, CD163, ADTRP, VWA8, G0S2, FCRL5, and Clorf226. Notably, when we performed qPCR validation in a new cohort of independent patients on seven of these genes, six of the seven (CYPIBI, ARMC12, OLAH, CD163, G0S2 and FCRL5) demonstrated consistent expression differences.

[0178] Biological role of classifier transcripts

[0179] Our data demonstrate that differential expression between patients with and without IA reflects neutrophil activation. Gene set enrichment analysis performed on the 95 differentially expressed transcripts (q<0.05) found in the training dataset showed

dysregulation of inflammatory and defense responses, and signaling (particularly IL-1, a major neutrophil activator) in neutrophils from patients with IA. These results mirror those of Example 1, which also showed differential expression was characteristic of increased peripheral activation. Indeed, several of the 26 classifier transcripts may be involved in neutrophil activation. CD 177 is a cell surface glycoprotein that plays a role in the activation and transmigration of neutrophils, while IL18R1 contributes to neutrophil activation through IL-18 signaling. Nectin-2 (PVRL2), a membrane glycoprotein, and PDE9A, a cGMP-specific phosphodiesterase, are both involved in cell adhesion. Additionally, lower PTGES expression may be partially responsible for increasing the lifespan of neutrophils, since it is involved in p53-induced apoptosis. Based on these findings, the neutrophil activation responses that may be dynamically involved in IA are reflected in our classification model, with aberrant expression of these transcripts increasing the likelihood of IA classification.

[0180] The differential expression of many other classifier transcripts reflects a complex inflammatory response in neutrophils in patients with IA. For example, CD 163 has been shown to be increased in neutrophils during sepsis and could be contributing to vascular inflammation during IA. Expression differences of other transcripts, like TGS1 and

CYP26B 1 (that are differentially expressed in tuberculosis, and juvenile idiopathic arthritis, respectively) could be related neutrophil functions that are pertinent to many responses to intravascular perturbations during IA. Other transcripts - such as ADTRP (expressed by macrophages in coronary artery plaques), OCLN (increased in activated T-lymphocytes and in whole blood during sepsis), and OLAH (increased in PBMCs during non-small cell lung cancer) - are involved in inflammation but have not been reported to be differentially expressed in neutrophils. Still, the roles of several other model transcripts (such as

MTRNR2L10, ARMC12, and LOC100506299) are widely unknown. Thus, it is believed the present disclosure provides the first demonstration of IA prediction from RNA expression patterns in the blood. [0181] Example 2, Table 1. Clinical characteristics.

Training Cohort Testing Cohort

Control (n=15) Aneurysm (n=15) Control (n=5) Aneurysm (n=5)

Age (MeaniSE) 59±4.8 63±2.8 63±7.2 52.6±6.6

Age [Median (Q1/Q3)] 61 (52.5/71.5) 64 (56.5/68.5) 68 (62/71) 53 (47/54)

Gender (number of patients)

Female 40% 66.67% 60% 40%

Smoking (number of patients)

Yes 0% 20% 40% 60%

Comorbidities (number of patients)

Hypertension 60% 60% 60% 20%

Heart Disease 6.67% 26.67% 40% 0%

High Cholesterol 26.67% 40% 60% 0%

Stroke History 6.67% 0% 0% 0%

Diabetes 33.33% 20% 20% 0%

Osteoarthritis 20% 33.33% 20% 0%

* Clinical characteristics of the randomly -created training and testing cohorts. With the exception of age, these factors were quantified as binary data points. The clinical factors were retrieved from the patients' medical records via the latest Patient Medical History form administered prior to imaging. Since this medical record contained self-reported information, the presence of each comorbidity was corroborated with each patients' reported list of medications (e.g. hypertension with Lisinopril, hyperhpidemia with simvastatin, heart disease with metoprolol, stroke history with clopidogrel, diabetes mellitus with metformin, and osteoarthritis with NSAIDs/tramadol). [0182] Example 2, Table 2. Gene ontology (GO) analysis.

Transcripts with higher expression in IA

Category GO Term Description p-value q-value

Process GO:0031347 Regulation of Defense Response 5.11E-06 0.0658

Process GO:0050727 Regulation of Inflammatory Response 1.01E-05 0.0652

Process GO:0019934 cGMP -Mediated Signaling 3.77E-05 0.162

Process GO:0032101 Regulation of Response to External 3.90E-05 0.125

Stimulus

Process GO:0031348 Negative Regulation of Defense Response 4.45E-05 0.115

Process GO:0050728 Negative Regulation of Inflammatory 5.21E-05 0.112

Response

Process GO:0007165 Signal Transduction 6.64E-05 0.122

Function GO:0004908 Interleukin-1 Receptor Activity 2.25E-06 0.00858 Function GO:0004872 Receptor Activity 7.22E-05 0.138

Function GO:0060089 Molecular Transducer Activity 7.22E-05 0.092

Function GO:0038023 Signaling Receptor Activity 1.32E-04 0.127 Transcripts with lower expression in IA

Category GO Term Description p-value q-value

Function GO:0043295 Glutathione Binding 1.16E-04 0.148

Function GO:0046906 Tetrapyrrole Binding 1.40E-04 0.134

* Gene set enrichment analysis was performed on the 95 significantly differentially expressed genes (q<0.05) peripheral blood samples obtained from patients with intracranial aneurysms (IA). Significantly enriched ontologies with an false discovery rate adjusted p-value (q-value)<0.20 were considered (FDR of 20%). Transcripts with higher expression in IA demonstrated regulation of inflammatory and defense responses, signaling and cell motility. Significantly enriched ontologies in transcripts with lower expression in IA demonstrated regulation of glutathione and tetrapyrrole binding.

Example 2, Table 3. 26 transcripts selected for classification model

G0S2 50486 NM_015714.3 -0.96 6.71E-06 2.66E-03

FCRL5 83416 NM 031281.2 -1.26 4.31E-06 1.80E-03

Clorf226 400793 NM_001135240.1 -1.51 1.27E-14 4.31E-11

UTS2 10911 NM_021995.2 -1.93 8.85E-14 1.66E-10

HBG2 3048 NM_000184.2 -1.97 6.62E-10 5.53E-07

CYP26B1 56603 NM_019885.3 -2.99 4.32E-07 2.37E-04

C1QL1 10882 NM_006688.4 -3.25 5.16E-22 3.88E-18

* Significantly differentially expressed transcripts with FDRO.05 and fold-change>1.5. Bolded transcripts additional to those 82 transcripts from Example .l(F-C=fold-change)

[0184] The cDNA nucleotide sequences of the 26 biomarkers in this table are provided as SEQ ID NOs 13-38, respectively.

[0185] Example 2, Supplemental Table 1. Batch Assignment.*

Batch (i.e. Example 1

ID Class or new pt in Example 2

Training Cohort

CI Control 1

C2 Control 1

C3 Control 1

C4 Control 1

C5 Control 1

C6 Control 1

C7 Control 1

C8 Control 2

C9 Control 2

CIO Control 2

Cl l Control 2

C12 Control 2

C13 Control 2

C14 Control 2

C15 Control 2

Al Aneurysm 1

A2 Aneurysm 1

A3 Aneurysm 1

A4 Aneurysm 1

A5 Aneurysm 1

A6 Aneurysm 1

A7 Aneurysm 1

A8 Aneurysm 1

A9 Aneurysm 2

A10 Aneurysm 2

Al l Aneurysm 2 A12 Aneurysm 2

A13 Aneurysm 2

A14 Aneurysm 2

A15 Aneurysm 1

Testing Cohort

C16 Control 1

C17 Control 1

C18 Control 1

C19 Control 2

C20 Control 1

A16 Aneurysm 1

A17 Aneurysm 1

A18 Aneurysm 2

A19 Aneurysm 2

A20 Aneurysm 2

*Over the study period to samples were collected and processed in two main batches. The earlier batch is designated as "1" and the later batch is designated as "2." When creating the training and testing cohorts, data from each batch was randomly partitioned into each group. [0186] Example 2, Supplemental Table 2. Primers used for qPCR and their efficiencies.*

*Primers were selected using Primer3 and NCBFs Primer Blast. All efficiencies were within the range of 0.90- 1.10. (bp=base pair, Eff=efficiency, Prod.=product, Temp.=temperature) le 2, Supplemental Table 3. RNA Quality. *

ID Class 260/280 RIN

Training Cohort

CI Control 2.02 7.4

C2 Control 1.92 6.5

C3 Control 2.08 7.1

C4 Control 2.05 7.9

C5 Control 1.96 6.7

C6 Control 2.08 6.2

C7 Control 2.04 7.3

C8 Control 2.04 6.6

C9 Control 2.04 6.0

CIO Control 2.05 6.0

Cl l Control 2.10 5.5

C12 Control 2.01 7.9

C13 Control 2.03 6.0

C14 Control 1.90 5.2

C15 Control 2.02 8.2

Al Aneurysm 2.04 7.8

A2 Aneurysm 2.07 7.5

A3 Aneurysm 2.02 6.1

A4 Aneurysm 2.03 7.3

A5 Aneurysm 1.99 6.5

A6 Aneurysm 2.02 7.5

A7 Aneurysm 2.05 7.7

A8 Aneurysm 1.95 7.2

A9 Aneurysm 2.07 6.7

A10 Aneurysm 2.00 6.9

Al l Aneurysm 1.97 6.4

A12 Aneurysm 2.12 7.3

A13 Aneurysm 1.96 7.4

A14 Aneurysm 2.07 7.8

A15 Aneurysm 1.97 6.9

Testing Cohort

C16 Control 2.07 7.0

C17 Control 2.05 6.6

C18 Control 1.99 7.1

C19 Control 1.92 6.6

C20 Control 1.97 6.4

A16 Aneurysm 2.06 8.1

A17 Aneurysm 2.07 6.0

A18 Aneurysm 2.08 7.1

A19 Aneurysm 2.03 7.2

A20 Aneurysm 2.06 6.0 * The quality of the RNA samples was assessed by the 260/280 ratio and the RIN. (RIN=RNA integrity number)

[0188] Example 2, Supplemental Table 4. Characteristics of 27 aneurysms in all patients with IAs (5 patients had multiple intracranial aneurysms)*

MRI for double vision

A17 19 mm ICA No No indicated possible IA

MRI for headache

A18 5.1 mm ICA No No indicated possible IA

Follow-up imaging of

A19 7 mm MCA +1 (3.5 mm ACA) No known IA

Follow-up imaging of

A20 2 mm ACA No No known IA

* Aneurysm size ranged from 1 mm to 19 mm. Sixteen IAs (60%) were classified as small (diameter <5 mm) and 11 (40%) were classified as large (diameter>5 mm). The aneurysms were situated at various locations in the Circle of Willis, with most being around the internal carotid artery (ICA) and its branches. Three patients with IAs had a family history of the disease. In general, digital subtraction angiography was performed for either confirmation of IA presence after an incidental finding of IA on noninvasive imaging, or for follow-up imaging of a previously detected IA. (ACA=anterior cerebral artery, AComA=anterior communicating artery, BT=basilar terminus, CT=computed tomography, DSA=digital subtraction angiography, IA=intracranial aneurysm, ICA=internal carotid artery, MCA=middle cerebral artery, MRA=magnetic resonance angiography, MRI=magnetic resonance imaging, PComA=posterior communicating artery, VB=vertebrobasilar)

[0189] Example 2, Supplemental Table 5. RNA Sequencing Quality Control

Analysis.*

Poor Seqs.

M. Qual. Leng % % M. Detected

ID Class Seqs. Seqs. th GC Aligned Aligned Transcripts

"raining Cohort

CI Control 59.3 0 51 49 96.50% 54.4 11434

C2 Control 68.9 0 51 49 96.60% 63.3 11928

C3 Control 80.3 0 51 50 96.30% 73.5 11829

C4 Control 97.4 0 51 50 96.00% 88.9 11064

C5 Control 67.8 0 51 49 96.40% 62.2 12135

C6 Control 36.2 0 51 49 96.10% 33.2 12057

C7 Control 79.3 0 51 49 96.80% 72.7 10117

C8 Control 66.0 0 51 50 95.50% 60.0 12048

C9 Control 74.4 0 51 50 95.10% 66.8 12154

CIO Control 64.3 0 51 51 94.90% 57.9 11866

Cl l Control 36.9 0 51 51 96.30% 33.7 11072

C12 Control 51.5 0 51 50 95.60% 46.8 11823

C13 Control 58.5 0 51 51 95.50% 53.0 11779

C14 Control 64.6 0 51 50 94.10% 57.7 11304

C15 Control 47.3 0 51 50 95.30% 42.7 12191

Al Aneurysm 48.1 0 51 49 96.90% 44.5 11656

A2 Aneurysm 35.7 0 51 48 96.80% 32.9 11101

A3 Aneurysm 55.9 0 51 49 96.90% 51.4 11725

A4 Aneurysm 61.3 0 51 49 97.20% 56.9 11087

A5 Aneurysm 23.4 0 51 49 95.90% 21.4 11951

A6 Aneurysm 35.8 0 51 49 96.40% 32.3 12108

A7 Aneurysm 26.9 0 51 50 97.10% 24.9 10151

A8 Aneurysm 59.5 0 51 49 96.00% 54.1 12111 A9 Aneurysm 86.9 0 51 51 84.60% 70.0 10809

A10 Aneurysm 64.9 0 51 50 96.60% 59.5 10869

Al l Aneurysm 75.3 0 51 51 95.90% 68.6 11938

A12 Aneurysm 47.2 0 51 51 89.50% 39.9 11512

A13 Aneurysm 42.1 0 51 50 95.20% 38.2 11185

A14 Aneurysm 32.3 0 51 50 88.50% 27.2 11903

A15 Aneurysm 14.6 0 51 49 95.00% 13.0 12127 resting Cohort

C16 Control 27 0 51 48 96.50% 24.8 11433

C17 Control 35.9 0 51 49 94.90% 32.3 12158

C18 Control 53.3 0 51 50 95.80% 48.5 11626

C19 Control 44.2 0 51 49 96.00% 40.6 12115

C20 Control 89.1 0 51 50 95.70% 81.1 11580

A16 Aneurysm 60.4 0 51 49 96.90% 55.9 12063

A17 Aneurysm 29.1 0 51 50 96.60% 26.8 11442

A18 Aneurysm 55.6 0 51 51 93.90% 49.3 10817

A19 Aneurysm 39.8 0 51 50 97.40% 36.8 11347

A20 Aneurysm 42.2 0 51 51 95.50% 38.3 12018

* The quality of the RNA sequencing experiments. Overall, prior to alignment all samples had an average of 53.84 M sequences. The sequencing experiments had an average of 48.4 M mapped reads with a 95.37% read mapping rate, and detected an average of 11591 transcripts (transcripts with TPM>l after batch effect correction). (M.=million, Seqs.=sequences, Qual.=quality)

[0190] Example 2, Supplemental Table 6. Clinical characteristics of the additional cohort of 5 patients with intracranial aneurysms and 5 control subjects without intracranial aneurysms (confirmed on imaging) used for qPCR validation.*

Patients with IA Control Patients

(n=5) without IA (n=5)

Age (Mean ± SE) 54.6±3.60 50.6±6.23

Age [Median (Q1/Q3)] 58 (46/60) 50 (39/62)

Gender

Female 40% 80%

Smoking Status

Yes 60% 0%

Comorbidities

Hypertension 0% 20%

High Cholesterol 60% 40%

Heart Disease 20% 0%

Stroke History 20% 40%

Diabetes 0% 20%

Osteoarthritis 20% 20% Clinical information, with the exception of age was quantified as binary data points. Clinical factors were retrieved from the patients' medical records via the latest Patient Medical History form administered prior to imaging, and the presence of each comorbidity was corroborated with each patients' reported list of medications. (IA=intracranial aneurysm, SE=standard error, Q=quartile)

[0191] Example 2, Supplemental Table 7. Characteristics of intracranial aneurysms in the group of 5 patients with IAs used for qPCR validation.*

* Aneurysm size ranged from 2.8 mm to 7 mm. Two of the 5 IAs (40%) were classified as small (<5mm) and three (60%) were classified as large (>5 mm). The aneurysms were situated in the Circle of Willis, in the anterior vasculature (MCA) and at the ICA. (CTA=computed tomography angiography, IA=intracranial aneurysm, ICA=internal carotid artery, MCA=middle cerebral artery, MRI=magnetic resonance imaging) [0192] Example 2, Supplemental Table 8. Summary of model performances across different analyses.*

Model training (LOO cross-validation)

Sen. Spec. Acc. AUC PPV* NPV*

Cosine NA ^T 0.87 0.60 0.73 0.72 0.10 0.99 n/ n. i ( i So o 73 1 ) 2 ⁽ 1 1 11

NSC 0.53 0.93 0.73 0.69 0.28 0.97

SVM 0.47 0.53 0.50 0.54 0.05 0.95

Model testing

Sen. Spec. Acc. AUC PPV* NPV*

Cosine NA ^r 0.60 0.60 0.60 0.62 0.07 0.97 / . i ⁽ I SO 1 no ο ι ι I I 8( 1 1 m i I ) w

NSC 0.80 0.80 0.80 0.76 0. 17 0.99

SVM 0.40 1.00 0.70 0.72 1.00 0.97 Sen. Spec. Acc. AUC PPV* NPV*

Cosine NN 0.75 0.65 0.70 0.70 0.10 0.98 n/ n. i ι κό ο ? ( 1 SO 1 ) S4 n 4 l 0.98

NSC 0.45 0.90 0.68 0.68 0. 19 0.97

SVM 0.40 0.85 0.63 0.69 0.12 0.96

* Model performances during training, testing and LOO retraining. Sensitivity, specific, accuracy, and AUC. *PPV and NPV calculated at 5% IA prevalence. Gray shading indicates best performing model.

(Sen.=sensitivity, Spec.=specificity, Acc.=accuracy, AUC=area under the receiver operating characteristic curve, PPV=positive predictive value, NPV=negative predictive value)

Example 3

[0193] This Example provides data demonstrating the use of biomarkers to discriminate large from small aneurisms. This is meaningful because the most widely-used metric for assessing the risk of an aneurysm rupturing is the aneurysms size. This size metric was adopted from longitudinal prospective studies which reported that larger aneurysms are more dangerous because they are more likely to rupture.

[0194] To determine if genes in circulating signatures of aneurysm could be used to assess the risk of aneurysm rupture, we assessed the capability of transcriptome profiling to identify significant differences between small and large intracranial aneurysms (small defined as <5mm, large as >5mm).

[0195] Study Population

[0196] This study was approved by the University at Buffalo Health Sciences

Institutional Review. Methods were carried out in accordance with the approved protocol. Written informed consent was obtained from all subjects. Patients undergoing cerebral digital subtraction angiography (DSA) with positive and negative intracranial aneurysm (IA) diagnoses were enrolled in this study. Reasons for the patients to receive DSA included confirmation of findings from noninvasive imaging of the presence of IAs, vascular malformations, or carotid stenosis, or follow-up noninvasive imaging of previously detected IAs. All consenting patients were older than 18 years, were English speaking, and had not received previous treatment for IA. We excluded patients who potentially had altered immune systems; this included patients who were pregnant, had recently undergone invasive surgery, were undergoing chemotherapy, had a body temperature above 37.78°C (100°F), had received solid organ transplants, had autoimmune diseases, and those who were taking prednisone or any other immunomodulating drugs. Furthermore, included patients did not have any other known cerebrovascular malformations or extracranial aneurysms, including abdominal aortic aneurysms. [0197] Neutrophil isolation

[0198] Sixteen mL of blood was drawn from the access catheter in the femoral artery and transferred into two 8 mL, citrated, cell preparation tubes (BD, Franklin Lakes, NJ). Neutrophils were isolated within 1 hour of peripheral blood collection. wwCell preparation tubes were centrifuged at 1,700 x g for 25 minutes to separate erythrocytes and neutrophils from mononuclear cells and plasma in the peripheral blood samples via a Ficoll density gradient. Erythrocytes and neutrophils were collected into a 3 mL syringe. Following hypotonic lysis of red blood cells, neutrophils were isolated by centrifugation at 400 x g for 10 min and disrupted and stored in TRIzol reagent (Life Technologies, Carlsbad, CA) at - 80°C until further processing. Neutrophils isolated in this fashion are more than 98%

CD66b+ by flow cytometry and contain no contaminating CD 14+ monocytes.

[0199] RNA preparation

[0200] Neutrophil RNA was extracted using TRIzol, according to the manufacturer's instructions. Trace DNA was removed by DNase I (Life Technologies, Carlsbad, CA) treatment. RNA was purified using the RNeasy MinElute Cleanup Kit (Qiagen, Venlo,

Limburg, Netherlands) and suspended in RNase-free water. The purity and concentration of RNA in each sample was measured by absorbance at 260 nm on a NanoDrop 2000 (Thermo Scientific, Waltham, MA), and 200-400 ng of RNA was sent to our university's Next- Generation Sequencing and Expression Analysis Core facility for further quality control. Precise RNA concentration was measured at the core facility via the Quant-iT RiboGreen Assay (Invitrogen, Carlsbad, CA) with a TBS-380 Fluorometer (Promega, Madison, WI), whereas the quality of the RNA samples was measured with an Agilent 2100 BioAnalyzer RNA 6000 Pico Chip (Agilent, Las Vegas, NV). RNA samples of acceptable purity (260/280 ratio of >1.9) and integrity (RIN > 5.0) were considered for RNA sequencing.

[0201] Results

[0202] The original cohort (as described in Example 1) contained 22 individuals (11 with IA, 11 without 11), from which we developed the 82-transcript signature. Subsequently, we added 18 patients (as described in Example 2), making our cohort 40 individuals (20 with IA, 20 without IA). Using this group, we developed the 26-transcript model. In this Example we added 26 more samples, resulting in a cohort of 66 samples (31 with IA, 35 without IA).

[0203] In the dataset of 31 samples from patients with aneurysms, there were 15 small aneurysm (<5 mm) samples and 16 large aneurysm (>5mm) samples. We applied the same statistical requirements described in Example 2 to evaluate differences in the transcriptomes of circulating neutrophils between control and aneurysm patients (TPM>1, q-value<0.05, and absolute fold-change>2). We found that 9 differentially expressed genes (ARMC12, Clorf226, CD177, OLAH, HRK, ITGA7, LYPD2, PTGDS, RPL39L) are present in the original 82-gene list developed in the 1 lvl 1 analysis (Example 1). Furthermore, 6 significantly differently expressed genes (ARMC12, Clorf226, CD177, HBG2,

LOCI 00506229, OLAH) were present in the 26-gene signature/model described in Example 2 (developed from 15vl5 analysis. These transcripts are reported in Example 3, Table 1 as "Gene Set 1". The transcripts bolded indicate they are present in both 82-gene signature and 26-gene model. With the exception of LYPD2 and PTGDS, the genes with increased expression IA from the signature and the model had even higher expression in larger aneurysms, and the genes with decreased expression in IA based on the signature and the model had even lower expression in larger aneurysms. This indicates that the present disclosure provides a method for generating a prediction of a size or size range of the aneurysm, and thus a risk of rupture. HRK, ITGA7, RPL39L were part of the original 82- gene set, but not part of the 26 gene panel, but still are significantly different between individuals with large and small aneurysms. Therefore, the disclosure encompasses determining these biomarkers. In certain embodiments, the disclosure encompasses including analysis of these biomarkers for use in differentiation of patients with large vs small IAs.

[0204] Example 3, Table 1: Gene Set 1

[0205] In this dataset we created a machine learning algorithm using a quadratic discriminant analysis model to relate the set of 11 IA-associated genes to the size of the IA (small<5mm, large>5mm). We used expression levels of these transcripts to train biomarker models using the machine learning toolbox in MATLAB. Five-fold cross-validation in this cohort estimated the model's accuracy to be 0.61 with an AUC of 0.61.

[0206] While the invention has been described through specific embodiments, routine modifications will be apparent to those skilled in the art and such modifications are intended to be within the scope of the present invention.