Methods for identifying subjects with Kawasaki disease

Abstract

A method for identifying a subject with Kawasaki disease (KD), the method comprising distinguishing the subject from a subject with another condition, such as other infectious and inflammatory conditions, such as conditions that exhibit symptoms similar to KD. Also provided are minimal gene signatures used in the method, as well as primers, probes, and gene chips used in the method.

Description

[0001] The present disclosure relates to a method for identifying a subject with Kawasaki disease (KD), the method comprising distinguishing the subject from a subject with another condition, such as other infectious and inflammatory conditions, such as a condition that exhibits symptoms similar to KD. The present disclosure also relates to a minimal gene signature employed in the method and a custom gene chip used in the method. The present disclosure further extends to probes and / or primers specific to the genes in the signature of the present disclosure. The present disclosure further relates to the use of known gene chips in the methods of the present disclosure and a kit comprising the elements required to perform the methods. The present disclosure also relates to the use of the method for providing a composite expression score, which can be used to distinguish bacterial infection from viral infection or inflammatory disease, particularly suitable for use in resource-poor environments. Background Art

[0002] Kawasaki disease (KD) is an acute inflammatory disorder that primarily affects young children. Since its initial description in Japan[1], KD has become the most common cause of acquired heart disease, with an incidence of 265 / 100,000 children under five years old in Japan[2], 51–194 / 100,000 in other Asian countries[3–4], and 8–20 / 100,000 in Europe[6] and the United States[7], respectively. Concerns about KD are related to its association with vasculitis, which primarily affects the coronary arteries, leading to the development of coronary artery aneurysms (CAAs) in up to 25% of untreated children[8]. Myocardial infarction may result from thrombotic occlusion of the aneurysm or from the development of stenotic lesions due to vascular remodeling in the damaged artery. Long-term outcome studies of children with giant CAAs have shown a worrying prognosis: within 30 years, more than 50% of patients require revascularization or suffer a myocardial infarction[9, 10].

[0003] Treatment with intravenous immunoglobulin (IVIG) and, in unresponsive patients, additional administration of IVIG

[11] or other anti-inflammatory agents (such as steroids and infliximab) can effectively abrogate the inflammatory process and reduce the risk of CAA by 5–10%

[12] . Because KD is difficult to distinguish from other common febrile conditions, many children with KD are not diagnosed and treated early enough in the course of their illness to prevent the development of CAA

[13] . In addition, patients who do not meet the clinical criteria for diagnostic KD (so-called “incomplete KD”) may still have CAA. Delayed diagnosis has always been a risk factor for the development of CAA, and even in centers with extensive experience in KD, treatment is often initiated only when coronary artery dilatation is already demonstrated on echocardiography. The development of CAA is clinically asymptomatic and may only be recognized years later with sudden death or myocardial infarction.

[0004] The symptoms of KD are similar to those of several other childhood febrile illnesses, including staphylococcal and streptococcal toxic shock syndrome, measles, and other viral illnesses (e.g., adenovirus infection, Rocky Mountain spotted fever, and childhood inflammatory disease), leading to diagnostic difficulties and, consequently, delays in diagnosis and treatment. Guidelines have been developed to facilitate clinical diagnosis based on clinical signs and symptoms, echocardiography, and laboratory parameters.

[14] However, there is no definitive diagnostic test for this disease. As the incidence of KD increases worldwide, there is an urgent need for an accurate test to differentiate KD from other conditions that cause prolonged fever in children. Summary of the invention

[0005] In the era of precision medicine, the diagnosis of many conditions that were previously based solely on clinical features has been replaced by molecular pathology-based diagnoses. Host blood gene expression signatures have been shown to distinguish a number of specific infectious and inflammatory diseases, including tuberculosis

[15] , bacterial and viral infections

[16] , and systemic lupus erythematosus

[17] . Support for gene expression signature-based approaches to KD diagnosis comes from the identification of microRNA biomarkers in KD [18, 19], although existing studies are limited in the scope of comparing patient groups or require RNA extraction from exosomes.

[0006] Therefore, the inventors have explored the use of whole blood gene expression patterns to distinguish KD from other childhood infectious and inflammatory conditions.The present disclosure provides a gene expression signature discovered and validated in an independent patient cohort that distinguishes KD from a range of bacterial, viral and inflammatory diseases.

[0007] The present disclosure is summarized in the following paragraphs:

[0008] 1. A method for identifying a subject having Kawasaki disease (KD), the method comprising detecting modulation of gene expression levels of a gene signature in an RNA sample derived from the subject, the gene signature comprising at least 5 of the following genes: CACNA1E, DDIAS, KLHL2, PYROXD2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163, and RTN1.

[0009] 2. The method of paragraph 1, wherein the gene signature comprises 6, 7, 8, 9, 10, 11, 12 or 13 of the genes.

[0010] 3. The method according to paragraph 1 or 2, wherein the gene signature comprises at least one of the following genes: PYROXD2, SMOX, CACNA1E, CD163, DDIAS, CLIC3, KLHL2 and HS.553068, specifically, at least one of PYROXD2, SMOX, CACNA1E and CD163.

[0011] 4. The method of any one of paragraphs 1 to 3, wherein the gene signature comprises PYROXD2.

[0012] 5. The method of any one of paragraphs 1 to 4, wherein the gene signature comprises CACNA1E.

[0013] 6. The method of any one of paragraphs 1 to 5, wherein the gene signature comprises SMOX.

[0014] 7. The method of any one of paragraphs 1 to 6, wherein the gene signature comprises CD163.

[0015] 8. The method of any one of paragraphs 1 to 7, wherein the gene signature comprises:

[0016] (i) PYROXD2 and CACNA1E;

[0017] (ii) PYROXD2 and SMOX; or

[0018] (iii) PYROXD2, CACNA1E and SMOX.

[0019] 9. The method according to any one of paragraphs 1 to 8, wherein the gene signature comprises or consists of at least 5 genes, for example, selected from the following genes:

[0020] (i) PYROXD2, CACNA1E, CD163, KLHL2, and SMOX;

[0021] (ii) PYROXD2, CACNA1E, IFI27, KLHL2, and SMOX;

[0022] (iii) PYROXD2, CACNA1E, HS.553068, IFI27, and SMOX;

[0023] (iv) PYROXD2, DDIAS, CACNA1E, IFI27, and SMOX; (v) PYROXD2, CACNA1E, CD163, KLHL2, and ZNF185; or

[0024] (vi) PYROXD2, DDIAS, CD163, KLHL2 and SMOX.

[0025] 10. The method according to any one of paragraphs 1 to 8, wherein the gene signature comprises or consists of at least 6 genes, for example, selected from the following genes:

[0026] (i) PYROXD2, CACNA1E, CD163, IFI27, KLHL2 and SMOX;

[0027] (ii) PYROXD2, CACNA1E, CD163, KLHL2, LINC02035, and SMOX;

[0028] (iii) PYROXD2, DDIAS, CACNA1E, CD163, IFI27, and SMOX;

[0029] (iv) PYROXD2, CACNA1E, CD163, HS.553068, IFI27, and SMOX;

[0030] (v) PYROXD2, CACNA1E, CD163, KLHL2, SMOX, and ZNF185;

[0031] (vi) PYROXD2, CACNA1E, IFI27, KLHL2, RTN1 and SMOX;

[0032] (vii) PYROXD2, CACNA1E, CD163, CLIC3, KLHL2, and SMOX;

[0033] (viii) PYROXD2, CACNA1E, CLIC3, IFI27, KLHL2, and SMOX;

[0034] (ix) PYROXD2, DDIAS, CACNA1E, IFI27, RTN1, and SMOX; or

[0035] (x) PYROXD2, DDIAS, CD163, IFI27, KLHL2 and SMOX.

[0036] 11. The method according to any one of paragraphs 1 to 8, wherein the gene signature comprises or consists of at least 7 genes, for example, selected from the following genes:

[0037] (i) PYROXD2, CACNA1E, CD163, HS.553068, IFI27, KLHL2, and SMOX;

[0038] (ii) PYROXD2, CACNA1E, CD163, CLIC3, IFI27, KLHL2, and SMOX;

[0039] (iii) PYROXD2, DDIAS, CACNA1E, CD163, IFI27, KLHL2, and SMOX;

[0040] (iv) PYROXD2, CACNA1E, CD163, IFI27, KLHL2, RTN1, and SMOX; or

[0041] (v) PYROXD2, DDIAS, CACNA1E, CD163, HS.553068, IFI27 and SMOX.

[0042] 12. The method according to any one of paragraphs 1 to 8, wherein the gene signature comprises or consists of at least 8 genes, for example, selected from the following genes:

[0043] (i) PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2 and SMOX;

[0044] (ii) PYROXD2, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, and SMOX;

[0045] (iii) PYROXD2, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1, and SMOX; or

[0046] (iv) PYROXD2, CACNA1E, CD163, HS.553068, IFI27, KLHL2, S100P and SMOX.

[0047] 13. The method according to any one of paragraphs 1 to 8, wherein the gene signature comprises or consists of at least 9 genes, for example, selected from the following genes:

[0048] (i) PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2 and SMOX;

[0049] (ii) PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1, and SMOX; or

[0050] (iii) PYROXD2, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1 and SMOX.

[0051] 14. The method according to any one of paragraphs 1 to 8, wherein the gene signature comprises or consists of at least 10 genes, for example, selected from the following genes:

[0052] (i) PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1, S100P, and SMOX; or

[0053] (ii) PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1 and SMOX.

[0054] 15. The method according to any one of paragraphs 1 to 8, wherein the gene signature comprises or consists of at least 11 genes, for example, selected from the following genes:

[0055] (i) PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1, S100P and SMOX;

[0056] (ii) PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, LINC02035, RTN1, S100P, and SMOX; or

[0057] (iii) PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1, S100P, SMOX, and ZNF185.

[0058] 16. The method according to any one of paragraphs 1 to 8, wherein the gene signature comprises or consists of at least 12 genes, for example, selected from the following genes:

[0059] (i) PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1, S100P, SMOX and ZNF185;

[0060] (ii) PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, LINC02035, RTN1, S100P, and SMOX; or

[0061] (iii) PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, LINC02035, RTN1, S100P, SMOX and ZNF185.

[0062] 17. The method of any one of paragraphs 1 to 8, wherein the gene signature comprises or consists of CACNA1E, DDIAS, KLHL2, PYROXD2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163, and RTN1.

[0063] 18. The method according to any one of paragraphs 1 to 17, wherein the method further comprises detecting the expression level or one or more housekeeping genes, such as 1, 2, 3, 4 or 5 housekeeping genes, for example, selected from actin, GAPDH, ubiquitin, 18s rRNA, RPII (POLR2A), TBP, PPIA, GUSB, HSPCB, YWHAZ, SDHA, RPS13, HPRT1 and B4GALT6.

[0064] 19. The method of any of paragraphs 1 to 18, wherein a subject with KD can be identified in the presence of one or more of: a bacterial infection, a viral infection, and an inflammatory condition.

[0065] 20. The method according to any of paragraphs 1 to 19, wherein subjects with KD can be distinguished from patients with one or more of: bacterial infections, viral infections, and inflammatory disorders.

[0066] 21. The method of paragraph 19 or 20, wherein the bacterial infection is selected from the group consisting of Chlamydia pneumoniae, Chlamydia trachomatis, Chlamydophila psittaci, Mycoplasma pneumonia, Corynebacterium diphtheriae, Clostridium botulinum, Clostridium difficile, Clostridium perfringens, Clostridium tetani, Enterococcus faecalis, Enterococcus faecium, Listeria monocytogenes, Staphylococcus aureus, Staphylococcus epidermidis, epidermidis, Staphylococcus saprophyticus, Group B streptococcus, Streptococcus agalactiae, Streptococcus pneumoniae, Streptococcus pyogenes, or acid-fast bacteria,Such as Mycobacterium leprae, Mycobaterium tuberculosis, Mycobacterium ulcerans, mycobacterium avium intercellularae, Bordetella pertussis, Borrelia burgdorferi, Brucella abortus, Brucella canis, Brucella melitensis, Brucella suis, Campylobacter jejuni, Escherichia coli, Francisella tularensis, Haemophilus influenzae, Helicobacter pylori, pylori), Legionella pneumophila, Leptospira interrogans, Neisseria gonorrhoeae, Neisseria meningitidis, Pseudomonas aeruginosa, Pseudomonas spp, Rickettsia rickettsii, Salmonella typhi, Salmonella typhimurium, Shigella sonnei, Treponemapallidum, Vibrio cholerae, Yersinia pestis, Kingella kingae), Stenotrophomonas, Klebsiella, Gram-positive cocci, Gram-negative bacilli, mycoplasmas, pertussis, mycobacteria, and staphylococcal and streptococcal toxic shock syndrome, e.g., Gram-positive cocci, Gram-negative bacilli, mycoplasmas or pertussis and mycobacteria, specifically,Selected from the group consisting of: Streptococcus pneumoniae, Staphylococcus aureus, Streptococcus pyogenes, Group B Streptococcus, Escherichia coli, Neisseria meningitidis, Enterococcus, Kingella, Haemophilus influenzae, Pseudomonas, Stenotrophomonas, Klebsiella, Staphylococcus and Streptococcus toxic shock syndrome, in particular, Staphylococcus or Streptococcus toxic shock syndrome. ,

[0067] 22. The method of any one of paragraphs 19 to 21, wherein the viral infection is selected from the group consisting of influenza, such as influenza A (including but not limited to: H1N1, H2N2, H3N2, H5N1, H7N7, H1N2, H9N2, H7N2, H7N3, H10N7), influenza B and influenza C, respiratory syncytial virus (RSV), rhinovirus, enterovirus, bocavirus, parainfluenza (such as parainfluenza types 1-4), adenovirus, metapneumovirus, herpes simplex virus, varicella virus, human papillomavirus, hepatitis, Epstein-Barr virus, varicella zoster virus, human cytomegalovirus, human herpes virus, BK virus type 8, JC virus, smallpox, parvovirus B19, human astrovirus, Norwalk virus, Coxsackie virus, Viruses, polioviruses, severe acute respiratory syndrome virus, yellow fever virus, dengue virus, West Nile virus, rubella virus, human immunodeficiency virus, Guanarito virus, Junin virus, Lassa virus, Machupo virus, Sabia virus, Crimean-Congo hemorrhagic fever virus, Ebola virus, Marburg virus, measles virus, mumps virus, rabies virus, rotavirus and Rocky Mountain spotted fever, for example, selected from the group consisting of respiratory syncytial virus (RSV), adenovirus, parainfluenza virus (such as parainfluenza types 1-4), influenza (such as influenza A, B or A+B), bocavirus, metapneumovirus, rhinovirus and enterovirus, in particular, RSV, influenza A / B and adenovirus, in particular, measles, adenovirus infection and Rocky Mountain spotted fever.

[0068] 23. The method according to any one of paragraphs 19 to 22, wherein the inflammatory disorder is selected from the group consisting of asthma, peptic ulcer, tuberculosis, periodontitis, ulcerative colitis, Crohn's disease, sinusitis, hepatitis, multiple sclerosis, atherosclerosis, Sjögren's disease, inflammatory bowel disease, lupus erythematosus (including systemic lupus erythematosus), fibrotic diseases (such as pulmonary fibrosis), Henoch-Schönlein purpura (HSP) and juvenile idiopathic arthritis (JIA), specifically, Henoch-Schönlein purpura (HSP) or juvenile idiopathic arthritis (JIA).

[0069] 24. The method of any one of paragraphs 1 to 23, wherein the subject is a child, e.g., the child is within the age range of 2 to 59 months.

[0070] 25. The method of any one of paragraphs 1 to 23, wherein the subject is an infant ranging in age from 0 to 59 days.

[0071] 26. The method of any one of paragraphs 1 to 25, wherein the subject has a fever.

[0072] 27. The method according to any one of paragraphs 1 to 26, wherein the analysis of gene expression regulation uses a microarray or a gene chip.

[0073] 28. The method according to any one of paragraphs 1 to 27, wherein the analysis of gene expression regulation employs: PCR, such as RT-PCR, in particular, multiplex PCR.

[0074] 29. The method of paragraph 14 or 15, wherein the PCR is quantitative.

[0075] 30. The method of any one of paragraphs 28 to 29, wherein the primers employed in the PCR comprise a label or a combination of labels, for example wherein the label is fluorescent or colored, such as colored beads.

[0076] 31. The method according to any one of paragraphs 1 to 30, comprising the further step of prescribing or administering a treatment for Kawasaki disease (KD) to the subject based on the analysis results of the genetic signature.

[0077] 32. A method for treating a subject having Kawasaki disease (KD), the method comprising administering a treatment for KD to the subject, wherein the subject has been previously identified as having Kawasaki disease by detecting modulation of gene expression levels of a gene signature in an RNA sample derived from the subject, the gene signature comprising at least 5 of the following genes: CACNA1E, DDIAS, KLHL2, PYROXD2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163, and RTN1.

[0078] 33. The method of paragraph 31 or 32, wherein the treatment is gamma globulin (IVIg), aspirin or other anti-inflammatory agents such as steroids and infliximab, or a combination thereof.

[0079] 34. A set of primers for use in a method for identifying a subject with Kawasaki disease (KD), the set of primers comprising primers specific for polynucleotide gene transcripts from at least 5 of the following genes: CACNA1E, DDIAS, KLHL2, PYROXD2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163, and RTN1.

[0080] 35. The primer set according to paragraph 34, which consists of primers specific only to the following genes: CACNA1E, DDIAS, KLHL2, PYROXD2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163 and RTN1.

[0081] 36. A gene chip consisting of probes specific to at least 5 of the following genes: CACNA1E, DDIAS, KLHL2, PYROXD2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163 and RTN1.

[0082] 37. A gene chip, consisting of: probes specific for at least 5 of the following genes: CACNA1E, DDIAS, KLHL2, PYROXD2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163 and RTN1; and one or more control probes.

[0083] 38. The gene chip according to paragraph 37, wherein the one or more control probes are specific to genes selected from the group consisting of actin, GAPDH, ubiquitin, 18s rRNA, RPII (POLR2A), TBP, PPIA, GUSB, HSPCB, YWHAZ, SDHA, RPS13, HPRT1 and B4GALT6.

[0084] 39. A point-of-care test for identifying a subject with Kawasaki disease (KD), the point-of-care test comprising a primer set as defined in paragraph 34 or 35 or a gene chip according to any one of paragraphs 36 to 38.

[0085] 40. Use of a primer set as defined in paragraph 34 or 35 or a gene chip according to any one of paragraphs 36 to 38 in an assay for detecting Kawasaki disease (KD) in a sample, such as a blood sample.

[0086] The present disclosure provides a method for identifying a subject with Kawasaki disease (KD), the method comprising detecting the expression level of at least 5 of the following genes: CACNA1E, DDIAS (C11ORF82), KLHL2, PYROXD2 (C10ORF33), SMOX, ZNF185, LINC02035 (LOC100129550), CLIC3, S100P, IFI27, HS.553068, CD163, and RTN1.

[0087] Thus, in one aspect, there is a method of identifying a subject having Kawasaki disease (KD), the method comprising detecting modulation of gene expression levels of a gene signature in an RNA sample derived from the subject, the gene signature comprising at least 5 of the following genes: CACNA1E, DDIAS (C11ORF82), KLHL2, PYROXD2 (C10ORF33), SMOX, ZNF185, LINC02035 (LOC100129550), CLIC3, S100P, IFI27, HS.553068, CD163, and RTN1.

[0088] Advantageously, the use of the gene signature in the methods according to the present disclosure allows for robust and accurate identification of subjects with KD. Importantly, the methods allow for accurate discrimination between patients with KD and patients who present with similar symptoms but suffer from other bacterial infections, viral infections, and / or inflammatory conditions. In other words, the methods allow for accurate detection of KD in the presence or absence of bacterial, viral infections, and / or inflammatory conditions without reliance on clinical criteria and / or laboratory tests, such as echocardiograms.

[0089] Gene signatures typically include a large number of genes that can only display biologically meaningful patterns or markers when combined together. It is very surprising that the gene signature of the present invention can be based on as few as 13 genes and still reliably identify the presence of KD. The gene signature of the present invention comprising at least 5 genes of the above-mentioned 13 genes provides good predictive power. However, additional genes may be included in the feature to further enhance and increase the discrimination of the gene signature. Therefore, in one embodiment, the feature comprises at least 5, 6, 7, 8, 9, 10, 11, 12 or 13 genes in the gene.

[0090] Thus, in one embodiment, the features include at least PYROXD2. In one embodiment, the features include at least SMOX. In one embodiment, the features include at least CACNA1E. In one embodiment, the features include at least CD163. In one embodiment, the features include at least DDIAS. In one embodiment, the features include at least CLIC3. In one embodiment, the features include at least KLHL2. In another embodiment, the features include at least HS.553068. In another embodiment, the features include at least RTN1. In another embodiment, the features include at least ZNF185. In another embodiment, the features include at least IFI27. In another embodiment, the features include at least S100P. In another embodiment, the features include at least LINC02035.

[0091] In one embodiment, the gene signature includes at least one of the following genes: PYROXD2, SMOX, CACNA1E, CD163, DDIAS, CLIC3, KLHL2, and HS.553068. In another embodiment, the gene signature includes at least one of the following genes: PYROXD2, SMOX, CACNA1E, and CD163. The inventors have found that these specific genes have higher discrimination and are therefore more likely to be present in the signature with the best predictive power.

[0092] For example, the gene signature may include or consist of any combination of the following genes: PYROXD2, SMOX, CACNA1E and CD163; PYROXD2, SMOX and CACNA1E; PYROXD2, SMOX and CD163; SMOX, CACNA1E and CD163; PYROXD2, CACNA1E and CD163; PYROXD2 and SMOX; PYROXD2 and CACNA1E; PYROXD2 and CD163; SMOX and CACNA1E; SMOX and CD163; or CACNA1E and CD163; or any other combination.

[0093] In one embodiment, the features include PYROXD2 and at least one of CACNA1E and SMOX. Thus, in one embodiment, the features include PYROXD2 and CACNA1E. In another embodiment, the features include PYROXD2 and SMOX. In yet another embodiment, the features include PYROXD2, CACNA1E, and SMOX.

[0094] In one embodiment, the feature includes at least 5 genes in the 13 genes. Therefore, in one embodiment, the feature includes or consists of: PYROXD2, CACNA1E, CD163, KLHL2 and SMOX. In one embodiment, the feature includes or consists of: PYROXD2, CACNA1E, IFI27, KLHL2 and SMOX. In one embodiment, the feature includes or consists of: PYROXD2, CACNA1E, HS.553068, IFI27 and SMOX. In one embodiment, the feature includes or consists of: PYROXD2, DDIAS, CACNA1E, IFI27 and SMOX. In one embodiment, the feature includes or consists of: PYROXD2, CACNA1E, CD163, KLHL2 and ZNF185. In one embodiment, the feature includes or consists of: PYROXD2, DDIAS, CD163, KLHL2 and SMOX.

[0095] In one embodiment, the feature includes at least 6 genes in the 13 genes. Therefore, in one embodiment, the feature includes or consists of: PYROXD2, CACNA1E, CD163, IFI27, KLHL2 and SMOX. In one embodiment, the feature includes or consists of: PYROXD2, CACNA1E, CD163, KLHL2, LINC02035 and SMOX. In one embodiment, the feature includes or consists of: PYROXD2, DDIAS, CACNA1E, CD163, IFI27 and SMOX. In one embodiment, the feature includes or consists of: PYROXD2, CACNA1E, CD163, HS.553068, IFI27 and SMOX. In one embodiment, the feature includes or consists of: PYROXD2, CACNA1E, CD163, KLHL2, SMOX and ZNF185. In one embodiment, the features include or consist of: PYROXD2, CACNA1E, IFI27, KLHL2, RTN1, and SMOX. In one embodiment, the features include or consist of: PYROXD2, CACNA1E, CD163, CLIC3, KLHL2, and SMOX. In one embodiment, the features include or consist of: PYROXD2, CACNA1E, CLIC3, IFI27, KLHL2, and SMOX. In one embodiment, the features include or consist of: PYROXD2, DDIAS, CACNA1E, IFI27, RTN1, and SMOX. In one embodiment, the features include or consist of: PYROXD2, DDIAS, CD163, IFI27, KLHL2, and SMOX.

[0096] In one embodiment, the feature includes at least 7 genes in the 13 genes. Therefore, in one embodiment, the feature includes the following or consists of: PYROXD2, CACNA1E, CD163, HS.553068, IFI27, KLHL2 and SMOX. In one embodiment, the feature includes the following or consists of: PYROXD2, CACNA1E, CD163, CLIC3, IFI27, KLHL2 and SMOX. In one embodiment, the feature includes the following or consists of: PYROXD2, DDIAS, CACNA1E, CD163, IFI27, KLHL2 and SMOX. In one embodiment, the feature includes the following or consists of: PYROXD2, CACNA1E, CD163, IFI27, KLHL2, RTN1 and SMOX. In one embodiment, the feature comprises or consists of PYROXD2, DDIAS, CACNA1E, CD163, HS.553068, IFI27, and SMOX.

[0097] In another embodiment, the feature includes at least 8 genes in the 13 genes. Therefore, in one embodiment, the feature includes the following or consists of: PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2 and SMOX. In another embodiment, the feature includes the following or consists of: PYROXD2, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2 and SMOX. In one embodiment, the feature includes the following or consists of: PYROXD2, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1 and SMOX. In one embodiment, the feature includes the following or consists of: PYROXD2, CACNA1E, CD163, HS.553068, IFI27, KLHL2, S100P and SMOX.

[0098] In another embodiment, the feature includes at least 9 genes in the 13 genes. Therefore, in one embodiment, the feature includes or consists of: PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2 and SMOX. In one embodiment, DDIAS, the feature includes or consists of: PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1 and SMOX. In one embodiment, the feature includes or consists of: PYROXD2, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1 and SMOX.

[0099] In another embodiment, the feature comprises at least 10 of the 13 genes. Thus, in one embodiment, the feature comprises or consists of: PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1, S100P, and SMOX. In another embodiment, the feature comprises or consists of: PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1, and SMOX.

[0100] In another embodiment, the feature includes at least 11 of the 13 genes. Therefore, in one embodiment, the feature includes or consists of: PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1, S100P and SMOX. In another embodiment, the feature includes or consists of: PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, LINC02035, RTN1, S100P and SMOX. In another embodiment, the feature includes or consists of: PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1, S100P, SMOX and ZNF185.

[0101] In one embodiment, the feature comprises at least 12 of the 13 genes. Thus, in one embodiment, the feature comprises or consists of: PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1, S100P, SMOX, and ZNF185. In another embodiment, the feature comprises or consists of: PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, LINC02035, RTN1, S100P, and SMOX. In another embodiment, the feature comprises or consists of PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, LINC02035, RTN1, S100P, SMOX, and ZNF185.

[0102] In one embodiment, the signature includes all 13 genes. Thus, in one embodiment, the gene signature includes or consists of: CACNA1E, DDIAS, KLHL2, PYROXD2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163, and RTN1. Advantageously, the signature including all 13 genes has the highest discrimination and allows for identification of KD with the highest degree of sensitivity and specificity.

[0103] Identification of KD is particularly important because KD is associated with vasculitis that may lead to the formation of coronary artery aneurysms (CCAs). Myocardial infarction may result from thrombotic occlusion of an aneurysm, or from the development of stenotic lesions due to vascular remodeling in the damaged artery. Therefore, there is a very large unmet clinical need for the correct and reliable identification of KD. The genetic signatures disclosed herein are a major step forward in the treatment of patients such as febrile patients, as the genetic signatures allow for accurate and rapid diagnosis, thereby allowing for appropriate and timely treatment of the patient.

[0104] Furthermore, the ingredients employed in the methods disclosed herein can be provided in a simple form that is cost-effective, rapid, economical, and can be used in low-resource and / or rural settings.

[0105] The present inventors found that the transcriptional expression levels of CACNA1E, DDIAS, KLHL2, PYROXD2, SMOX, ZNF185, LINC02035 and CLIC3 were increased in subjects with KD compared to subjects without KD, and the transcriptional expression levels of S100P, IFI27, HS.553068, CD163 and RTN1 were decreased in subjects with KD compared to subjects without KD.

[0106] Advantageously, the inventors were able to use a gene signature that detects modulation of gene expression levels of the 13 genes listed above to discriminate subjects with KD from subjects without KD, which has High AUC and high sensitivity and specificity

[0107] Advantageously, the gene signature is developed from a training set containing a range of ethnicities. This means that the gene signature and method disclosed herein can be applied to samples from subjects of different ethnicities. Further advantageously, the gene signature was developed using KD patients who were ill for no more than 7 days. This means that the signature can help in early diagnosis of KD before the onset of fever for 5 days, which helps in early identification of KD patients and allows for appropriate treatment early on.

[0108] Thus, the inventors have demonstrated that the method is applicable to a wide range of different samples and patient groups, indicating that the method is robust and reliable.

[0109] Thus, in one aspect, the present disclosure provides a method for diagnosing a subject with Kawasaki disease, the method comprising detecting modulation of gene expression levels of a gene signature in an RNA sample derived from the subject, the gene signature comprising at least 5 of the following genes: CACNA1E, DDIAS (C11ORF82), KLHL2, PYROXD2 (C10ORF33), SMOX, ZNF185, LINC02035 (LOC100129550), CLIC3, S100P, IFI27, HS.553068, CD163, and RTN1.

[0110] In one embodiment, the diagnostic method is performed in vitro.

[0111] In one embodiment, the method further employs one or more housekeeping genes, such as 1, 2, 3, 4 or 5 housekeeping genes. In the context of this specification, housekeeping genes are not considered to be part of the feature. In one embodiment, the housekeeping gene is selected from the group consisting of: actin, GAPDH, ubiquitin, 18s rRNA, RPII (POLR2A), TBP, PPIA, GUSB, HSPCB, YWHAZ, SDHA, RPS13, HPRT1 and B4GALT6.

[0112] In one embodiment, the methods of the present disclosure are capable of identifying a subject suffering from KD in the presence of a bacterial infection, a viral infection, and / or an inflammatory condition.

[0113] In one embodiment, the methods of the present disclosure are capable of distinguishing subjects suffering from KD from patients suffering from bacterial infections, viral infections, and / or inflammatory disorders.

[0114] In one embodiment, the bacterial infection is selected from the group consisting of: Chlamydia pneumoniae, Chlamydia trachomatis, Chlamydia psittaci, Mycoplasma pneumoniae, Corynebacterium diphtheriae, Clostridium botulinum, Clostridium difficile, Clostridium perfringens, Clostridium tetani, Enterococcus faecalis, Enterococcus faecium, Listeria monocytogenes, Staphylococcus aureus, Staphylococcus epidermidis, Staphylococcus saprophyticus, Group B Streptococcus, Streptococcus agalactiae, Streptococcus pneumoniae, Streptococcus pyogenes, or acid-fast bacteria such as Mycobacterium leprae, Mycobacterium tuberculosis, Mycobacterium ulcerans, Mycobacterium avium, Bordetella pertussis, Borrelia burgdorferi, Brucella abortus, Brucella canis, Brucella melitensis, Brucella suis bacteria, Campylobacter jejuni, Escherichia coli, Francisella tularensis, Haemophilus influenzae, Helicobacter pylori, Legionella pneumophila, Leptospira interrogans, Neisseria gonorrhoeae, Neisseria meningitidis, Pseudomonas aeruginosa, Pseudomonas, Rickettsia rickettsii, Salmonella typhi, Salmonella typhimurium, Shigella sonnei, Treponema pallidum, Vibrio cholerae, Yersinia pestis, Kingella kingae, Stenotrophomonas, Klebsiella, Gram-positive cocci, Gram-negative bacilli, Mycoplasma, Pertussis, Mycobacterium, and Staphylococcus and Streptococcus toxic shock syndrome, such as Gram-positive cocci, Gram-negative bacilli, Mycoplasma, or Pertussis, and Mycobacterium.

[0115] In one embodiment, the bacterial infection is selected from the group consisting of: Streptococcus pneumoniae, Staphylococcus aureus, Streptococcus pyogenes, Group B Streptococcus, Escherichia coli, Neisseria meningitidis, Enterococcus, Kingella, Haemophilus influenzae, Pseudomonas, Stenotrophomonas, and Klebsiella.

[0116] In one embodiment, the bacterial infection is Staphylococcus or Streptococcus toxic shock syndrome.

[0117] In one embodiment, the viral infection is selected from the group consisting of: influenza, such as influenza A (including but not limited to: H1N1, H2N2, H3N2, H5N1, H7N7, H1N2, H9N2, H7N2, H7N3, H10N7), influenza B and influenza C, respiratory syncytial virus (RSV), rhinovirus, enterovirus, bocavirus, parainfluenza, adenovirus, metapneumovirus, herpes simplex virus, varicella virus, human papillomavirus, hepatitis, Epstein-Barr virus, varicella zoster virus, human Cytomegalovirus, human herpesvirus, BK virus type 8, JC virus, smallpox, parvovirus B19, human astrovirus, Norwalk virus, coxsackievirus, poliovirus, severe acute respiratory syndrome virus, yellow fever virus, dengue virus, West Nile virus, rubella virus, human immunodeficiency virus, Guanarito virus, Junin virus, Lassa virus, Machupo virus, Sabia virus, Crimean-Congo hemorrhagic fever virus, Ebola virus, Marburg virus, measles virus, mumps virus, rabies virus, rotavirus, and Rocky Mountain spotted fever.

[0118] In one embodiment, the viral infection is selected from the group consisting of: respiratory syncytial virus (RSV), adenovirus, parainfluenza virus (such as parainfluenza type 1-4), influenza (such as influenza type A, type B or influenza type A+B), bocavirus, metapneumovirus, rhinovirus and enterovirus, specifically, RSV, influenza type A / type B and adenovirus. In one embodiment, the viral infection is selected from the group consisting of: measles, adenovirus infection and Rocky Mountain spotted fever.

[0119] The method according to any one of claims 4 to 11, wherein the inflammatory disorder is selected from the group consisting of asthma, peptic ulcer, tuberculosis, periodontitis, ulcerative colitis, Crohn's disease, sinusitis, hepatitis, multiple sclerosis, atherosclerosis, Sjögren's disease, inflammatory bowel disease, lupus erythematosus (including systemic lupus erythematosus), fibrotic diseases (such as pulmonary fibrosis), Henoch-Schonlein purpura (HSP) and juvenile idiopathic arthritis (JIA).

[0120] In one embodiment, the inflammatory disease is the following: juvenile idiopathic arthritis (JIA), Henoch-Schönlein purpura (HSP).

[0121] In another aspect, the present disclosure provides a method of treating a subject diagnosed with KD using the methods herein.

[0122] In one embodiment, the subject is a child, eg, under 17 years of age, such as 2 to 59 months old.

[0123] In one embodiment, the subject is an infant, eg, within the age range of 0 to 59 days.

[0124] In one embodiment, the subject has a fever, such as a febrile patient.

[0125] In one embodiment, the methods of the present disclosure are used with patient-derived samples, such as blood samples.

[0126] In one embodiment, analysis of gene expression modulation employs a microarray.

[0127] In one embodiment, analysis of gene expression modulation employs PCR, such as RT-PCR.

[0128] In one embodiment, the PCR is a multiplex PCR.

[0129] In one embodiment, the PCR is quantitative.

[0130] In one embodiment, the primers used in the PCR include a label or a combination of labels.

[0131] In one embodiment, the label is fluorescent or colored, for example the label is a colored bead.

[0132] In one embodiment, analysis of gene expression modulation employs two-color reverse transcriptase multiplex ligation-dependent probe amplification.

[0133] In one embodiment, the gene expression modulation is detected by using fluorescence spectroscopy.

[0134] In one embodiment, the modulation of gene expression is detected by using a colorimetric assay.

[0135] In one embodiment, said gene expression modulation is detected by using impedance spectroscopy.

[0136] In one embodiment, the method comprises the further step of prescribing or administering a treatment to the subject suffering from KD based on the results of the analysis of the genetic signature.

[0137] Thus, in one aspect, a method of treating a KD patient by administering a treatment such as gamma globulin (IVIg), aspirin or other anti-inflammatory agents such as steroids and infliximab is provided, wherein the patient is characterized in that the patient has been identified as KD positive by the methods disclosed herein. Thus, in one aspect, a method of treating a subject with Kawasaki disease (KD) is provided, the method comprising administering a treatment for KD to the subject, wherein the subject has previously been identified as having Kawasaki disease by detecting modulation of gene expression levels of a gene signature in an RNA sample derived from the subject, the gene signature comprising at least 5 of the following genes: CACNA1E, DDIAS, KLHL2, PYROXD2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163 and RTN1. Suitable treatments for KD are known to the skilled artisan and include, but are not limited to, gamma globulin (IVIg), aspirin or other anti-inflammatory agents, such as steroids and infliximab, or combinations thereof.

[0138] In one aspect, a method of determining whether to administer a KD treatment such as gamma globulin (IVIg), aspirin, or other anti-inflammatory agents (such as steroids and infliximab) is provided, the method comprising the steps of performing a method according to the present disclosure, and if the method indicates that the subject has KD, administering the KD to the subject.

[0139] Thus, the presently disclosed methods can help to appropriately treat patients such as patients with fever, for example, in cases where it is unclear whether the fever is due to Kawasaki disease or to a bacterial infection, a viral infection, an inflammatory condition, or a combination thereof. This has the advantage of ensuring rapid and appropriate treatment without having to wait for laboratory test results.

[0140] In one aspect of the present disclosure, a primer set for multiplex PCR is provided, wherein the primer set comprises a nucleic acid sequence specific to a polynucleotide gene transcript of at least 5 genes from the following genes: CACNA1E, DDIAS, KLHL2, PYROXD2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163 and RTN1. Therefore, in one embodiment, the primer set has specificity at least from the transcript of PYROXD2. In one embodiment, the primer set has specificity at least from the transcript of SMOX. In one embodiment, the primer set has specificity at least from the transcript of CACNA1E. In one embodiment, the primer set has specificity at least from the transcript of CD163. In one embodiment, the primer set has specificity at least from the transcript of DDIAS. In one embodiment, the primer set has specificity at least from the transcript of CLIC3. In one embodiment, the primer set is specific for transcripts from at least KLHL2. In one embodiment, the primer set is specific for transcripts from at least HS.553068. In one embodiment, the primer set is specific for transcripts from at least RTN1. In one embodiment, the primer set is specific for transcripts from at least ZNF185. In one embodiment, the primer set is specific for transcripts from at least IFI27. In one embodiment, the primer set is specific for transcripts from at least S100P. In one embodiment, the primer set is specific for transcripts from at least LINC02035.

[0141] In one embodiment, the primer set is specific to a transcript of at least one gene from the following genes: PYROXD2, SMOX, CACNA1E, CD163, DDIAS, CLIC3, KLHL2, and HS.553068. In another embodiment, the primer set is specific to a transcript of at least one gene from the following genes: PYROXD2, SMOX, CACNA1E, and CD163.

[0142] For example, the primer set is specific for transcripts from any combination of the following genes: PYROXD2, SMOX, CACNA1E, and CD163; PYROXD2, SMOX, and CACNA1E; PYROXD2, SMOX, and CD163; SMOX, CACNA1E, and CD163; PYROXD2, CACNA1E, and CD163; PYROXD2 and SMOX; PYROXD2 and CACNA1E; PYROXD2 and CD163; SMOX and CACNA1E; SMOX and CD163; or CACNA1E and CD163; or any other combination.

[0143] In one embodiment, the primer set is specific for transcripts from PYROXD2 and at least one of CACNA1E and SMOX. Thus, in one embodiment, the primer set is specific for PYROXD2 and CACNA1E. In another embodiment, the primer set is specific for PYROXD2 and SMOX. In yet another embodiment, the primer set is specific for PYROXD2, CACNA1E and SMOX.

[0144] In one embodiment, the primer set is specific for transcripts from at least 5 genes in the 13 genes: Therefore, in one embodiment, the primer set is specific for PYROXD2, CACNA1E, CD163, KLHL2 and SMOX. In one embodiment, the primer set is specific for PYROXD2, CACNA1E, IFI27, KLHL2 and SMOX. In one embodiment, the primer set is specific for PYROXD2, CACNA1E, HS.553068, IFI27 and SMOX. In one embodiment, the primer set is specific for PYROXD2, DDIAS, CACNA1E, IFI27 and SMOX. In one embodiment, the primer set is specific for PYROXD2, CACNA1E, CD163, KLHL2 and ZNF185. In one embodiment, the primer set is specific for PYROXD2, DDIAS, CD163, KLHL2 and SMOX.

[0145] In one embodiment, the primer set is specific for at least 6 genes among the 13 genes. Therefore, in one embodiment, the primer set is specific for PYROXD2, CACNA1E, CD163, IFI27, KLHL2 and SMOX. In one embodiment, the primer set is specific for PYROXD2, CACNA1E, CD163, KLHL2, LINC02035 and SMOX. In one embodiment, the primer set is specific for PYROXD2, DDIAS, CACNA1E, CD163, IFI27 and SMOX. In one embodiment, the primer set is specific for PYROXD2, CACNA1E, CD163, HS.553068, IFI27 and SMOX. In one embodiment, the primer set is specific for PYROXD2, CACNA1E, CD163, KLHL2, SMOX and ZNF185. In one embodiment, the primer set is specific for PYROXD2, CACNA1E, IFI27, KLHL2, RTN1 and SMOX. In one embodiment, the primer set is specific for PYROXD2, CACNA1E, CD163, CLIC3, KLHL2 and SMOX. In one embodiment, the primer set is specific for PYROXD2, CACNA1E, CLIC3, IFI27, KLHL2 and SMOX. In one embodiment, the primer set is specific for PYROXD2, DDIAS, CACNA1E, IFI27, RTN1 and SMOX. In one embodiment, the primer set is specific for PYROXD2, DDIAS, CD163, IFI27, KLHL2 and SMOX.

[0146] In one embodiment, the feature includes at least 7 genes in the 13 genes.Therefore, in one embodiment, the primer set is specific to PYROXD2, CACNA1E, CD163, HS.553068, IFI27, KLHL2 and SMOX. In one embodiment, the primer set is specific to PYROXD2, CACNA1E, CD163, CLIC3, IFI27, KLHL2 and SMOX. In one embodiment, DDIAS, the primer set is specific to PYROXD2, DDIAS, CACNA1E, CD163, IFI27, KLHL2 and SMOX. In one embodiment, the primer set is specific to PYROXD2, CACNA1E, CD163, IFI27, KLHL2 and SMOX. In one embodiment, DDIAS, the primer set is specific for PYROXD2, DDIAS, CACNA1E, CD163, HS.553068, IFI27, and SMOX.

[0147] In another embodiment, the primer set is specific to at least 8 genes in the 13 genes. Therefore, in one embodiment, the primer set is specific to PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2 and SMOX. In another embodiment, the primer set is specific to PYROXD2, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2 and SMOX. In one embodiment, the primer set is specific to PYROXD2, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1 and SMOX. In one embodiment, the primer set is specific to PYROXD2, CACNA1E, CD163, HS.553068, IFI27, KLHL2, S100P and SMOX.

[0148] In another embodiment, the primer set is specific for at least 9 genes in the 13 genes. Therefore, in one embodiment, the primer set is specific for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2 and SMOX. In one embodiment, the primer set is specific for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1 and SMOX. In one embodiment, the primer set is specific for PYROXD2, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1 and SMOX.

[0149] In another embodiment, the primer set is specific for at least 10 of the 13 genes. Thus, in one embodiment, the primer set is specific for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1, S100P, and SMOX. In another embodiment, the primer set is specific for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1, and SMOX.

[0150] In another embodiment, the primer set is specific for at least 11 of the 13 genes. Thus, in one embodiment, the primer set is specific for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1, S100P, and SMOX. In another embodiment, the primer set is specific for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, LINC02035, RTN1, S100P, and SMOX. In another embodiment, the primer set is specific for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1, S100P, and SMOX.

[0151] In one embodiment, the primer set is specific for at least 12 of the 13 genes. Thus, in one embodiment, the primer set is specific for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1, S100P, SMOX, and ZNF185. In another embodiment, the primer set is specific for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, LINC02035, RTN1, S100P, and SMOX. In another embodiment, the primer set is specific for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, LINC02035, RTN1, S100P, SMOX, and ZNF185.

[0152] In one embodiment, the primer set is specific for all 13 genes. Thus, in one embodiment, the primer set is specific for CACNA1E, DDIAS, KLHL2, PYROXD2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163, and RTN1.

[0153] In one embodiment, the gene transcript is RNA, such as mRNA or cRNA.

[0154] In one embodiment, the primers for each gene are at least a pair of nucleic acid primer sequences.

[0155] In one embodiment, the length of the primers is 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 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 200, 201

[0156] In one embodiment, at least one primer for each gene includes a tag.

[0157] In one embodiment, the labels on the primers are independently selected from fluorescent labels, color labels, antibodies, step tags, and his tags.

[0158] In one embodiment, each primer in a given primer pair is labeled, for example wherein one label quenches the fluorescence of the other label when the labels are adjacent to each other.

[0159] In another aspect of the present disclosure, a gene chip is provided, which is composed of probes for detecting the regulation of gene expression levels of at least 5 genes among the following genes: CACNA1E, DDIAS, KLHL2, PYROXD2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163 and RTN1.

[0160] In one embodiment, the Illumina probe IDs for the 13 genes are shown in Table 2. Alternatively, a skilled artisan can design custom probes based on the nucleic acid sequence of each of the 13 genes.

[0161] In one embodiment, the gene chip further includes a control probe. In the context of the present disclosure, the control probe is not considered to be a part of the gene feature. Therefore, in one embodiment, the gene chip is composed of: probes for at least 5 genes in the following genes: CACNA1E, DDIAS, KLHL2, PYROXD2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163 and RTN1; and one or more control probes. In one embodiment, the control probe is specific to the transcripts of one or more genes from the following genes: actin, GAPDH, ubiquitin, 18s rRNA, RPII (POLR2A), TBP, PPIA, GUSB, HSPCB, YWHAZ, SDHA, RPS13, HPRT1 and B4GALT6.

[0162] Advantageously, a chip with probes for at least 5 of the 13 genes can accurately and reliably distinguish samples, such as whole blood from subjects with KD from samples from subjects with bacterial / viral infections and / or inflammatory conditions. Such a chip with a small number of probes can be produced inexpensively, making the chip particularly suitable for use in resource-poor settings.

[0163] Therefore, in one embodiment, the gene chip includes at least a probe for PYROXD2. In one embodiment, the gene chip includes at least a probe for SMOX. In one embodiment, the gene chip includes at least a probe for CACNA1E. In one embodiment, the gene chip includes at least a probe for CD163. In one embodiment, the gene chip includes at least a probe for DDIAS. In one embodiment, the gene chip includes at least a probe for CLIC3. In one embodiment, the gene chip includes at least a probe for KLHL2. In one embodiment, the gene chip includes at least a probe for HS.553068. In one embodiment, the gene chip includes at least a probe for RTN1. In one embodiment, the gene chip includes at least a probe for ZNF185. In one embodiment, the gene chip includes at least a probe for IFI27. In one embodiment, the gene chip includes at least a probe for S100P. In one embodiment, the gene chip includes at least a probe for LINC02035.

[0164] In one embodiment, the gene chip includes probes for at least one of the following genes: PYROXD2, SMOX, CACNA1E, CD163, DDIAS, CLIC3, KLHL2, and HS.553068. In another embodiment, the gene chip includes probes for at least one of the following genes: PYROXD2, SMOX, CACNA1E, and CD163.

[0165] For example, the gene chip includes probes for any of the following gene combinations: PYROXD2, SMOX, CACNA1E and CD163; PYROXD2, SMOX and CACNA1E; PYROXD2, SMOX and CD163; SMOX, CACNA1E and CD163; PYROXD2, CACNA1E and CD163; PYROXD2 and SMOX; PYROXD2 and CACNA1E; PYROXD2 and CD163; SMOX and CACNA1E; SMOX and CD163; or CACNA1E and CD163; or any other combination.

[0166] In one embodiment, the gene chip includes probes for PYROXD2 and at least one of CACNA1E and SMOX. Therefore, in one embodiment, the gene chip includes probes for PYROXD2 and CACNA1E. In another embodiment, the gene chip includes probes for PYROXD2 and SMOX. In yet another embodiment, the gene chip includes probes for PYROXD2, CACNA1E and SMOX.

[0167] In one embodiment, the gene chip includes probes for at least 5 genes in the 13 genes or consists of the probes. Therefore, in one embodiment, the gene chip includes probes for PYROXD2, CACNA1E, CD163, KLHL2 and SMOX or consists of the probes. In one embodiment, the gene chip includes probes for PYROXD2, CACNA1E, IFI27, KLHL2 and SMOX or consists of the probes. In one embodiment, the gene chip includes probes for PYROXD2, CACNA1E, HS.553068, IFI27 and SMOX or consists of the probes. In one embodiment, the gene chip includes probes for PYROXD2, DDIAS, CACNA1E, IFI27 and SMOX or consists of the probes. In one embodiment, the gene chip includes probes for PYROXD2, CACNA1E, CD163, KLHL2 and ZNF185 or consists of the probes. In one embodiment, the gene chip comprises or consists of probes for PYROXD2, DDIAS, CD163, KLHL2 and SMOX.

[0168] In one embodiment, the gene chip includes probes for at least 6 of the 13 genes or consists of the probes. Therefore, in one embodiment, the gene chip includes probes for PYROXD2, CACNA1E, CD163, IFI27, KLHL2 and SMOX or consists of the probes. In one embodiment, the gene chip includes probes for PYROXD2, CACNA1E, CD163, KLHL2, LINC02035 and SMOX or consists of the probes. In one embodiment, the gene chip includes probes for PYROXD2, DDIAS, CACNA1E, CD163, IFI27 and SMOX or consists of the probes. In one embodiment, the gene chip includes probes for PYROXD2, CACNA1E, CD163, HS.553068, IFI27 and SMOX or consists of the probes. In one embodiment, the gene chip comprises or consists of probes for PYROXD2, CACNA1E, CD163, KLHL2, SMOX and ZNF185. In one embodiment, the gene chip comprises or consists of probes for PYROXD2, CACNA1E, IFI27, KLHL2, RTN1 and SMOX. In one embodiment, the gene chip comprises or consists of probes for PYROXD2, CACNA1E, CD163, CLIC3, KLHL2 and SMOX. In one embodiment, the gene chip comprises or consists of probes for PYROXD2, CACNA1E, CLIC3, IFI27, KLHL2 and SMOX. In one embodiment, the gene chip comprises or consists of probes for PYROXD2, DDIAS, CACNA1E, IFI27, RTN1 and SMOX. In one embodiment, the gene chip comprises or consists of probes for PYROXD2, DDIAS, CD163, IFI27, KLHL2 and SMOX.

[0169] In one embodiment, the gene chip includes probes for at least 7 of the 13 genes or consists of the probes. Therefore, in one embodiment, the gene chip includes probes for PYROXD2, CACNA1E, CD163, HS.553068, IFI27, KLHL2 and SMOX or consists of the probes. In one embodiment, the gene chip includes probes for PYROXD2, CACNA1E, CD163, CLIC3, IFI27, KLHL2 and SMOX or consists of the probes. In one embodiment, DDIAS, the gene chip includes probes for PYROXD2, DDIAS, CACNA1E, CD163, IFI27, KLHL2 and SMOX or consists of the probes. In one embodiment, the gene chip includes probes for PYROXD2, CACNA1E, CD163, IFI27, KLHL2 and SMOX or consists of the probes. In one embodiment, the gene chip includes probes for PYROXD2, CACNA1E, CD163, IFI27, KLHL2, RTN1 and SMOX or consists of the probes. In one embodiment, the gene chip comprises or consists of probes for PYROXD2, DDIAS, CACNA1E, CD163, HS.553068, IFI27 and SMOX.

[0170] In another embodiment, the gene chip comprises probes for at least 8 of the 13 genes or consists of the probes. Therefore, in one embodiment, the gene chip comprises probes for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2 and SMOX or consists of the probes. In another embodiment, the gene chip comprises probes for PYROXD2, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2 and SMOX or consists of the probes. In one embodiment, the gene chip comprises probes for PYROXD2, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1 and SMOX or consists of the probes. In one embodiment, the gene chip comprises or consists of probes for PYROXD2, CACNA1E, CD163, HS.553068, IFI27, KLHL2, S100P and SMOX.

[0171] In another embodiment, the gene chip comprises probes for at least 9 of the 13 genes or consists of the probes. Therefore, in one embodiment, the gene chip comprises probes for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2 and SMOX or consists of the probes. In one embodiment, the gene chip comprises probes for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1 and SMOX or consists of the probes. In one embodiment, the gene chip comprises probes for PYROXD2, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1 and SMOX or consists of the probes.

[0172] In another embodiment, the gene chip comprises probes for at least 10 of the 13 genes or consists of the probes. Therefore, in one embodiment, the gene chip comprises probes for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1, S100P and SMOX or consists of the probes. In another embodiment, the gene chip comprises probes for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1 and SMOX or consists of the probes.

[0173] In another embodiment, the gene chip comprises probes for at least 11 of the 13 genes or consists of the probes. Therefore, in one embodiment, the gene chip comprises probes for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1, S100P and SMOX or consists of the probes. In another embodiment, the gene chip comprises probes for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, LINC02035, RTN1, S100P and SMOX or consists of the probes. In another embodiment, the gene chip comprises or consists of probes for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1, S100P, SMOX and ZNF185.

[0174] In one embodiment, the gene chip comprises probes for at least 12 of the 13 genes or consists of the probes. Therefore, in one embodiment, the gene chip comprises probes for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1, S100P, SMOX and ZNF185 or consists of the probes. In another embodiment, the gene chip comprises probes for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, LINC02035, RTN1, S100P and SMOX or consists of the probes. In another embodiment, the gene chip comprises or consists of probes for PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, LINC02035, RTN1, S100P, SMOX and ZNF185.

[0175] In one embodiment, the gene chip includes probes for all 13 genes or is composed of the probes. Therefore, in one embodiment, the gene chip includes probes for CACNA1E, DDIAS, KLHL2, PYROXD2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163 and RTN1 or is composed of the probes.

[0176] In further embodiments, the present disclosure encompasses the use of a known or commercially available gene chip in the methods of the present disclosure.

[0177] In one aspect, a point-of-care test for identifying a subject with KD is provided, the point-of-care test comprising a primer set or a gene chip as defined above. Advantageously, the currently disclosed test can be rapidly performed in just a few hours without the need for complex diagnostic or laboratory equipment. Therefore, the currently disclosed method can be easily implemented in a hospital setting as well as in a resource-poor setting (such as in a remote village) as part of an existing patient care program.

[0178] In one aspect, there is provided a use of a primer set or a gene chip as defined above in an assay for detecting KD in a sample, such as a blood sample. DETAILED DESCRIPTION

[0179] The 13 genes / gene transcripts shown in Table 2 can be used to identify patients with KD or distinguish KD from bacterial infection. In one embodiment, the method of the present disclosure is able to distinguish subjects with KD from different conditions / diseases or infections (such as bacterial / viral infections or inflammatory conditions) with similar clinical symptoms. In another embodiment, these 13 genes / gene transcripts can be used to distinguish viral infections. In yet another embodiment, these 13 genes / gene transcripts can be used to distinguish patients with KD from inflammatory diseases such as juvenile idiopathic arthritis (JIA), Henoch-Schönlein purpura (HSP) or systemic lupus erythematosus (SLE).

[0180] In one embodiment, one probe is used to detect modulation of gene expression for each gene, for example selected from the list of probes shown in Table 2.

[0181] In another embodiment, two or more probes are used to detect regulation of each gene.In one embodiment of the present disclosure, the gene signature is the minimum set of genes required to optimally detect infection or discriminate disease (eg, between bacterial / viral infections and / or between inflammatory diseases).

[0182] Optimal refers to the minimal set of genes that discriminates between KD and bacterial / viral infection and / or inflammatory conditions without significant loss of specificity and / or sensitivity of characteristic detection or discrimination capabilities.

[0183] As used herein, detection is intended to refer to the process of identifying KD in a sample, specifically, by detecting the modulation of relevant genes in a signature. In one embodiment, it may be detected that the subject only has KD. In another embodiment, the subject may have KD and also have a bacterial infection, a viral infection, an inflammatory condition, or a combination thereof.

[0184] Discrimination refers to the ability of a signature to differentiate between different disease states (e.g., KD versus viral / bacterial infection or inflammatory disease). In the context of gene signatures, detection and discrimination are interchangeable.

[0185] As used herein, a subject is a person suspected of having KD or a person with a fever from whom the sample is obtained. The term patient may be used interchangeably, although in one embodiment the patient has a condition.

[0186] In one embodiment, the methods of the disclosure are performed on a sample from a subject having or suspected of having KD, eg, where the subject exhibits symptoms typically associated with KD.

[0187] In one embodiment, the methods of the present disclosure are performed on samples from subjects who have or are suspected of having a bacterial / viral infection or an inflammatory condition but are not suspected of having KD, e.g., where the subject exhibits symptoms not typically associated with KD. Testing samples from such subjects can help identify individuals with KD who would not normally be correctly diagnosed.

[0188] In one embodiment, the subject exhibits symptoms of a viral infection. In another embodiment, the subject exhibits symptoms of a bacterial infection. In yet another embodiment, the subject exhibits symptoms of both a bacterial infection and a viral infection. In one embodiment, the subject exhibits symptoms of an inflammatory condition.

[0189] In further embodiments, the sample is a sample from a febrile subject; that is, a temperature above the normal body temperature of 37.5°C.

[0190] In yet further embodiments, an analysis is performed to determine whether the fever is associated with KD. Determining the source of the fever / infection advantageously allows for the prescription and / or administration of appropriate medications, for example, appropriate treatments such as gamma globulin (IVIg), aspirin, etc., can be administered to patients with KD, while antibiotics can be administered to patients with bacterial infections, and antipyretics can be administered to patients with viral infections.

[0191] Efficient treatment is beneficial because it minimizes hospital stays, ensures that patients receive appropriate treatment, which can save lives, especially when the patient is an infant or child, and also ensures that resources are used wisely.

[0192] In recent years, it has become clear that overuse of antibiotics should be avoided because it can lead to bacterial resistance. Therefore, the use of antibiotics in patients who do not have a bacterial infection should be avoided.

[0193] In one embodiment, the subject is an adult. An adult is defined herein as a person over 18 years of age.

[0194] In one embodiment, the subject is a child. As used herein, a child refers to a person under the age of 18 (eg, 5 to 17 years old).

[0195] In one embodiment, the subject is an infant. As used herein, an infant refers to a human within the age range of 0 to 59 days.

[0196] As used herein, modulation of gene expression refers to the upregulation or downregulation of one or more genes.

[0197] As used herein, up-regulated is intended to refer to gene transcripts that are expressed at higher levels in disease or infection patient samples relative to, for example, control samples that do not contain the relevant disease or infection, or samples having a potential disease or infection or at a different stage of a disease or infection, as the case may be.

[0198] As used herein, down-regulated is intended to refer to gene transcripts that are expressed at lower levels in samples of patients with a disease or infection relative to, for example, control samples that do not contain the relevant disease or infection, or samples with a potential disease or infection or at a different stage of a disease or infection. Thus, an up-regulated gene is a gene that is expressed at a higher level in a subject with KD compared to a subject without KD. Likewise, a down-regulated gene is expressed at a lower level in a subject with KD compared to a subject without KD.

[0199] Regulation is measured by measuring the level of gene expression using appropriate techniques.

[0200] As used herein, gene expression is the process of using information from genes for the synthesis of functional gene products. These products are usually proteins, but in non-protein coding genes (such as ribosomal RNA (rRNA), transfer RNA (tRNA) or small nuclear RNA (snRNA) genes), the products are functional RNA. In other words, RNA has a function. In the context of the present disclosure, measuring the expression level of a gene generally refers to measuring the level of a transcript associated with the gene.

[0201] As employed herein, gene expression data is intended to refer to any data generated from a patient sample that indicates the expression of two or more (e.g., 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, or 50) genes.

[0202] In one embodiment, one or more (e.g., 1 to 21, such as 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 or 20) genes are replaced by functionally equivalent genes, provided that the signature retains the ability to detect / distinguish the relevant clinical state without a significant decrease in specificity and / or sensitivity.

[0203] In one embodiment, the genes used are identical to the 13 genes listed in Table 2.

[0204] In one embodiment, one or more genes in the 13 gene signature are significantly differentially expressed in samples from subjects with KD compared to samples from subjects without KD.

[0205] As used herein, a gene signature is intended to refer to two or more genes that, when tested together, are capable of detecting / distinguishing a related clinical state. Thus, a gene signature represents a minimal set of genes that has sufficient discriminatory power to identify subjects with KD or to distinguish subjects with KD from subjects with bacterial / viral infection or inflammatory disease.

[0206] As used herein, significantly differentially expressed means that a gene shows a log2 fold change >0.5 or <-0.5 in samples from subjects with KD compared to samples from subjects not with KD (e.g., subjects with bacterial / viral infection and / or inflammatory disorder).

[0207] In one embodiment, as used herein, up-regulated means that the gene shows a log2 fold change > 0.5.

[0208] In one embodiment, as used herein, down-regulated means that the gene shows a log2 fold change < -0.5.

[0209] In one embodiment, one or more of the following genes are downregulated in a subject with KD: S100P, IFI27, HS.553068, CD163, and RTN1.

[0210] In one embodiment, one or more of the following genes are upregulated in subjects with KD: CACNA1E, DDIAS, KLHL2, PYROXD2, SMOX, ZNF185, LINC02035, CLIC3.

[0211] As used herein, present in the described form refers to placing genes from one or more features in the form of probes on a microarray.

[0212] As used herein, accurate and robust refers to the fact that the method can be employed in real-world or low-resource environments (such as Africa) and the results of the method's execution are correct giving a high confidence that the real results are obtained.

[0213] The method will provide high confidence when the method provides few false positive results (e.g., the results indicate that the subject has a bacterial infection when the subject does not have a bacterial infection) and also few false negatives (e.g., the results indicate that the subject does not have a bacterial infection when the subject does have a bacterial infection).

[0214] When an appropriate statistical test is employed, high confidence would include 90% or higher confidence, such as 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% confidence.

[0215] In one embodiment, the method provides a sensitivity of 80% or higher (such as 90% or higher, specifically 95% or higher), for example, wherein the sensitivity is calculated as follows:

[0216]

[0217] In one embodiment, the method provides a high level of specificity, in particular, for example, 80% or higher, such as 90% or higher, in particular, 95% or higher, for example, wherein the specificity is calculated as follows:

[0218]

[0219] In one embodiment, the sensitivity of the gene signature method is 90% to 100%, such as 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%.

[0220] In one embodiment, the specificity of the gene signature method is 85% to 100%, such as 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%.

[0221] In one embodiment, the sensitivity of the gene signature method is 85% to 100%, such as 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%.

[0222] In one embodiment, the specificity of the gene signature method is 85% to 100%, such as 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%.

[0223] There are a variety of methods by which gene expression can be measured, including microarrays, tiling arrays, DNA or RNA arrays (eg, on gene chips), RNA-seq, and serial analysis of gene expression.

[0224] Any suitable method of measuring gene regulation can be employed in the methods of the present disclosure.

[0225] In one embodiment, the gene expression measured is that of a host (eg, a human), such as an inflammatory response of the host, ie, not that of an infectious agent or disease.

[0226] In one embodiment, DNA or RNA from a sample from a subject is analyzed.

[0227] In one embodiment, RNA from a sample from a subject is analyzed.

[0228] In one embodiment, mRNA from a sample from a subject is analyzed.

[0229] In one embodiment, cRNA from a sample from a subject is analyzed.

[0230] In one embodiment, the sample is a solid or a fluid, such as blood or serum or a processed form of either.

[0231] As used herein, a fluid sample refers to a liquid derived from a living human body. The liquid includes liquids excreted or secreted from the body, and water that is not usually excreted from the body. It includes amniotic fluid, aqueous humor and vitreous humor, bile, serum, breast milk, cerebrospinal fluid, cerumen (earwax), chyle, endolymph and perilymph, gastric juice, mucus (including nasal drainage and phlegm), sputum, peritoneal fluid, pleural fluid, saliva, sebum (skin oil), semen, sweat, tears, vaginal secretions, vomitus, urine. Specifically, blood and serum.

[0232] As used herein, blood refers to whole blood, ie, serum, blood cells and coagulation factors, typically peripheral whole blood.

[0233] As used herein, serum refers to the components of whole blood other than blood cells or coagulation factors. It is plasma from which fibrinogen has been removed.

[0234] In one embodiment, the subject-derived sample is a blood sample.

[0235] In one embodiment, the sample is whole blood. Thus, in one embodiment, the RNA sample is from whole blood.

[0236] The RNA sample can be further amplified by PCR, such as whole genome amplification, to increase the amount of starting RNA template available for analysis. Alternatively, the RNA sample can be converted to cDNA by a reverse transcriptase such as HIV-1 reverse transcriptase, Moloney murine leukemia virus (M-MLV) reverse transcriptase, AMV reverse transcriptase, and telomerase reverse transcriptase. This amplification step may be required for smaller sample volumes, such as blood samples obtained from children.

[0237] In one or more embodiments, the analysis is ex vivo.

[0238] As used herein, ex vivo means occurring outside the body.

[0239] In one embodiment, the gene expression data is generated from a microarray (eg, a gene chip).

[0240] As used herein, microarrays include RNA or DNA arrays, such as RNA arrays. The skilled person will be aware of various different forms of microarrays, including but not limited to solid phase arrays and microbead arrays.

[0241] As used herein, polymerase chain reaction (PCR) refers to a widely used molecular technique for preparing multiple copies of a target DNA sequence. The method relies on thermal cycling, which consists of a cycle of repeated heating and cooling reactions to perform DNA melting and enzymatic replication of the DNA. The method is named after a primer containing a sequence complementary to the target region and a DNA polymerase, which is a key component for achieving selectivity and repeated amplification. As PCR proceeds, the generated DNA itself will be used as a replication template, thereby initiating a chain reaction in which the DNA template is exponentially amplified.

[0242] As used herein, multiplex PCR refers to the use of the polymerase chain reaction (PCR) to amplify two or more different DNA sequences simultaneously, ie, as if many separate PCR reactions were performed together in one reaction.

[0243] As used herein, primer is intended to refer to a short nucleic acid sequence, usually a chemically synthesized oligonucleotide, which serves as the starting point of a DNA synthesis reaction.

[0244] Primers are usually about 15 base pairs in length, but can vary in length from 5 to 100 bases. This is necessary in processes such as PCR because DNA polymerase can only add new nucleotides, or base pairs, to an existing DNA strand. During the PCR reaction, the primers hybridize to complementary sequences in the DNA sample. Next, the DNA polymerase begins replication at the 3' end of the primer and extends the primer by copying the sequence of the opposite DNA strand.

[0245] In one embodiment, the primers disclosed herein are specific to RNA (such as mRNA), i.e., the primers are complementary to the RNA sequence. In another embodiment, the primers are specific to cDNA, i.e., the primers are complementary to the cDNA sequence.

[0246] In one embodiment, the primers of the present disclosure include a label that enables the primer to be detected or isolated. Examples of labels include but are not limited to fluorescent labels, color labels, and antibodies, step tags, and his tags.

[0247] In another embodiment, each primer in a given primer pair is labeled, for example, wherein when the labels are adjacent to each other, a label (also referred to as a quencher) quenches the fluorescence of another label. For example, such labels are particularly useful in real-time PCR reactions. Examples of such labels include 6-carboxyfluorescein (FAM) and tetrachlorofluorescein, or tetramethylrhodamine and tetrachlorofluorescein.

[0248] As used herein, point-of-care testing or bedside testing is intended to refer to medical diagnostic tests performed at or near the point of care (i.e., at the time and place of patient care). This is in contrast to conventional diagnostic tests, which are typically limited to medical laboratories and involve sending specimens from the point of care to the laboratory for testing. Such diagnostic tests typically require hours or days to receive test results. At the same time, patient care must continue without knowing the test results. In contrast, point-of-care testing is typically a simple medical test that can be performed quickly.

[0249] A gene chip is essentially a microarray, that is, an array of discrete areas (usually nucleic acids) separated from each other, for example at about 100 / cm 2 Up to 1000 / cm 2 density, but can be arranged at a higher density (such as 10000 / cm 2 )arrangement.

[0250] The principle of a microarray experiment is to use mRNA from a given cell line or tissue to generate a labeled sample, usually a labeled cDNA or cRNA, which is called a "target", which is hybridized in parallel with a large number of nucleic acid sequences (usually DNA or RNA sequences) fixed on a solid surface. Hundreds of thousands of transcripts can be detected and quantified simultaneously. Although many different microarray systems have been developed, the most commonly used systems today can be divided into two groups.

[0251] Using this technique, arrays consisting of more than 30,000 cDNAs can be assembled onto the surface of a conventional microscope slide. For oligonucleotide arrays, short 20-25mers can be synthesized in situ by photolithography on silicon wafers (Affymetrix's High Density Oligonucleotide Arrays) or by inkjet technology (developed by Rosetta Inpharmatics and licensed by Agilent Technologies).

[0252] Alternatively, pre-synthesized oligonucleotides can be printed on a glass slide. The method based on synthetic oligonucleotides has the following advantages: since the sequence information alone is sufficient to generate the DNA to be arrayed, time-consuming cDNA resource processing is not required. Similarly, probes can be designed to represent the most unique part of a given transcript, so that closely related genes or splice variants can be detected. Although short oligonucleotides may lead to less specific hybridization and reduce sensitivity, arrays of pre-synthesized longer oligonucleotides (50-100 polymers) have recently been developed to compensate for these shortcomings.

[0253] In one embodiment, the gene chip is a commercially available chip, such as the human HT-12v4 expression BeadChip kit available from Illumina, the NimbleGen microarray available from Roche, Agilent, and Eppendorf, and the gene chip available from Affymetrix, such as the HU-Ul 33.Plus 2.0 gene chip.

[0254] In another embodiment, the gene chip used in the present invention is a customized gene chip, that is, the chip contains only target genes associated with the desired spectrum. Customized chips can be purchased from companies such as Roche, Affymetrix, etc. In yet another embodiment, the customized gene chip includes a minimal set of disease-specific transcripts.

[0255] In one embodiment, the chip consists of probes for detecting the expression levels of the 13 genes listed in Table 2.

[0256] In one embodiment, the following Illumina probe ID numbers are used to detect modulation of gene expression levels: 7510647 for CACNA1E, 2570019 for DDIAS, 1070593 for KLHL2, 1684497 for PYROXD2, 270068 or 3710553 for SMOX, 6840674 for ZNF185, 3236239 for LINC02035, 5870136 for CLIC3, 1510424 for S100P, 3990170 for IFI27, 1470450 for HS.553068, 2680092 for CD163, and 6860193 for RTN1.

[0257] In one or more of the above embodiments, the chip may further comprise one or more (eg, 1 to 10) control probes, such as housekeeping genes.

[0258] In one embodiment, gene expression data is generated in solution using appropriate probes for the genes of interest.

[0259] As used herein, probe is intended to refer to a hybridization probe, which is a fragment of DNA or RNA of variable length (usually 100-1000 bases long) that is used in a DNA or RNA sample to detect the presence of a nucleotide sequence (DNA target) that is complementary to the sequence in the probe. The probe thus hybridizes with a single-stranded nucleic acid (DNA or RNA) whose base sequence allows probe-target base pairing due to the complementarity between the probe and the target.

[0260] In one embodiment, the methods according to the present disclosure and, for example, the chips used therein may include one or more housekeeping genes.

[0261] As used herein, housekeeping genes are intended to refer to genes that are not directly relevant to a profile for identifying a disease or infection but that can be used for statistical purposes and / or quality control purposes, such as to help normalize the data, specifically, housekeeping genes are constitutive genes, i.e., genes that are transcribed at a relatively constant level. The products of housekeeping genes are generally required for the maintenance of cells.

[0262] Examples of housekeeping genes include, but are not limited to, actin, GAPDH, ubiquitin, 18s rRNA, RPII (POLR2A), TBP, PPIA, GUSB, HSPCB, YWHAZ, SDHA, RPS13, HPRT1, and B4GALT6.

[0263] In one embodiment, as used herein, a minimal disease-specific transcript set refers to the minimum number of genes required to robustly identify a disease state of interest.

[0264] The minimal discriminatory gene set can be interchanged with the minimal disease-specific transcript set or minimal gene signature.

[0265] As used herein, normalization is intended to refer to statistically accounting for background noise by comparing the data to control data (e.g., fluorescence levels of a housekeeping gene), e.g., fluorescence scan data can be normalized using RMA to allow comparisons between individual chips. This method was described by Irizarry et al. in 2003.

[0266] As used herein, scaling refers to enhancing the contribution of a specific gene that is expressed at a low level or has a high fold change but still has a relatively low fluorescence, thereby increasing the contribution of the specific gene to the diagnostic feature.

[0267] Fold change is commonly used in gene expression data analysis in microarray and RNA-Seq experiments to measure changes in gene expression levels and is simply calculated as the ratio of the final value to the initial value, i.e. if the initial value is A and the final value is B, the fold change is B / A (Tusher et al., 2001).

[0268] In programs such as Arrayminer, fold changes in gene expression can be calculated. Statistics associated with fold changes are calculated, and the statistics are more significant in genes with less variation in expression levels between different groups of subjects (e.g., in genes with greater differences between groups).

[0269] The step of obtaining a suitable sample from the subject is a conventional technique involving the collection of a blood sample. This process poses little risk to the donor and need not be performed by a physician, but may be performed by appropriately trained support staff. In one embodiment, the sample from the subject is approximately 2.5 ml of blood, but smaller volumes, such as 0.5-1 ml, may be used.

[0270] Immediately place blood or other tissue fluid in RNA stabilization buffer, such as that contained in Pax Gene Tubes or Tempus Tubes.

[0271] If storage is required, it should usually be frozen at -80°C within 3 hours of collection.

[0272] In one embodiment, gene expression data is generated from RNA levels in a sample.

[0273] For microarray analysis, blood can be processed using a suitable product such as the PAX Gene Blood RNA Extraction Kit (Qiagen).

[0274] Total RNA can also be purified using the Tripure method - Tripure extraction (Roche catalog number 1667165). The manufacturer's protocol may be followed. Purification can then be performed using the RNeasy Mini kit - clean-up protocol with DNAse treatment (Qiagen catalog number 74106).

[0275] Quantification of RNA can be accomplished using optical density at 260 nm and Quant-IT RiboGreen RNA Assay Kit (Invitrogen-Molecular Probes R1 1490). The quality of the 28s and 18s ribosomal RNA peaks can be assessed using an Agilent Bioanalyzer.

[0276] In another embodiment, the method further comprises the step of amplifying the RNA.Amplification can be performed using a suitable kit, such as the TotalPrep RNA Amplification Kit (Applied Biosystems).

[0277] In one embodiment, the amplification method can be combined with labeling of RNA for microarray analysis. Nugen 3'ovation Biotin Kit (Cat. Nos.: 2300-12, 2300-60).

[0278] RNA from the subject sample is then hybridized to relevant probes, such as those that may be located on a chip. Following hybridization and washing, analysis is performed using appropriate instrumentation, where appropriate.

[0279] When performing an analysis to determine whether a subject according to the present disclosure exhibits a genetic signature indicative of a disease or infection, the following steps are performed: obtaining mRNA from a sample and preparing nucleic acid targets; hybridizing to an array under appropriate conditions, typically as recommended by the microarray manufacturer (appropriately stringent hybridization conditions, such as 3X SSC, 0.1% SDS at 50°C), to bind to the corresponding probes on the array; and washing, if necessary, to remove unbound nucleic acid targets and analyzing the results.

[0280] In one embodiment, the readout from the assay is fluorescence.

[0281] In one embodiment, the readout from the assay is colorimetric.

[0282] In one embodiment, physical detection methods may be used, such as changes in electrical impedance, nanowire technology, or microfluidics technology.

[0283] In one embodiment, a method is provided, the method further comprising the step of quantifying RNA from a sample from a subject.

[0284] If a quality control step is required, software such as Genome Studio software can be used.

[0285] As used herein, a numerical value refers to the amount of each gene of interest obtained from an analysis or readout of gene expression (e.g., fluorescence or colorimetric analysis). The numerical value obtained from the initial analysis can be manipulated, corrected, and if the result of the processing is still a number, it will continue to be a numerical value.

[0286] Conversion refers to the process of treating a negative value as positive or a positive value as negative by simply converting the positive sign to a negative sign or vice versa.

[0287] Analysis of a sample derived from a subject will give a series of numerical values ​​for the genes analyzed, some of which are positive (beginning with +, which is considered mathematically greater than zero) and others are negative (beginning with –, which is considered strictly mathematically less than zero). Positive and negative in gene expression analysis are a convenient mechanism for representing upregulated genes and downregulated genes.

[0288] In the method of the present disclosure, all numerical values ​​of genes that are downregulated and represented by negative numbers will be converted to corresponding positive numbers (i.e., by simply changing the sign), for example -1 will be converted to 1; or all positive numerical values ​​of genes that are upregulated will be converted to corresponding negative numbers.

[0289] The inventors have determined that the step of making the numerical values ​​for gene expression all positive or all negative allows the values ​​to be summed to obtain a single value indicative of the presence or absence of a disease or infection.

[0290] This greatly simplifies the processing of gene expression data and represents a practical advance, making the method suitable for routine clinical use.

[0291] Discrimination refers to the ability to distinguish KD samples from bacterially infected samples, virally infected samples / subjects and / or between and inflammatory diseases such as SLE, JIA and HSP.

[0292] For example, the discriminatory power of methods according to the present disclosure can be increased by attaching greater weights to genes that are more important in a signature, even if they are expressed at low or lower absolute levels.

[0293] As used herein, raw values ​​are intended to refer to unprocessed fluorescence values, for example from a gene chip, either absolute fluorescence or relative to one or more housekeeping genes.

[0294] As employed herein, summing is intended to refer to the act or process of adding numerical values.

[0295] As used herein, a composite expression score refers to the sum (total) of all individual values ​​generated by analysis for relevant genes, e.g., the sum of fluorescence data for all relevant up-regulated and down-regulated genes. Scores may be normalized and / or unnormalized and / or scaled and / or weighted.

[0296] In one embodiment, the composite expression score is normalized.

[0297] In one embodiment, the composite expression score is scaled.

[0298] In one embodiment, the composite expression score is weighted.

[0299] As employed herein, weighted or statistical weighting is intended to refer to a relevant value being adjusted to more appropriately reflect its contribution to a feature.

[0300] In one embodiment, the method employs a simplified risk score as employed in the examples herein.

[0301] The simplified risk score is also called the disease risk score (DRS).

[0302] As used herein, control is intended to refer to positive (control) samples and / or negative (control) samples that are used, for example, to compare a subject sample to and / or a defined value or range of values ​​to allow the subject sample to be designated as positive or negative for a disease / infection by its reference.

[0303] As used herein, a positive control sample is a sample known to be positive for the pathogen or disease (eg, bacterial infection) being analyzed.

[0304] As used herein, a negative control sample is intended to refer to a sample that is known to be negative for the pathogen or disease being analyzed.

[0305] In one embodiment, the control is a sample, such as a positive control sample or a negative control sample, such as a negative control sample.

[0306] In one embodiment, a control is a value, such as a range of values, such as a statistically determined range obtained from a sufficient sample size, that defines a cutoff value for accurately distinguishing disease cases from controls.

[0307] Converting a multi-gene transcript disease signature into a single disease score

[0308] Once the RNA expression signature of the disease has been determined through variable selection, the transcripts can be separated based on whether they are up- or down-regulated relative to the comparison group. Two groups of transcripts are selected and collated separately.

[0309] Summing up-regulated and down-regulated RNA transcripts

[0310] In order to determine a single disease risk score for any single patient, the original intensities of all upregulated RNA transcripts associated with the disease, such as fluorescence intensity (absolute intensity or relative to the housekeeping standard), are summed. By combining the original values ​​(such as fluorescence) of each transcript relative to the unchanged housekeeping gene standard, a similar summation of all downregulated transcripts for each individual can be achieved. Since transcripts have different expression levels and their multiple changes are also different, it is not necessary to sum the original expression values, and they can be scaled and normalized between 0 and 1. Alternatively, it can be weighted so that important genes play a greater role. Then, for each sample, the expression values ​​of the transcripts of the features are summed for the upregulated and downregulated transcripts, respectively.

[0311] An overall disease score containing the cumulative fluorescence of up- and down-regulated genes was calculated by adding the cumulative score of down-regulated transcripts (after conversion to a positive number) to the cumulative score of up-regulated transcripts to give a single-number composite expression score. This score maximally distinguishes cases from controls and reflects the contribution of up- and down-regulated transcripts to this distinction.

[0312] Comparison of disease risk scores in cases and controls

[0313] Composite expression scores for patients and controls can be compared to derive means and variances for each group, from which statistical cutoffs are defined to accurately distinguish cases from controls. Using disease subjects and comparison populations, sensitivity and specificity of disease risk scores can be calculated using, for example, support vector machines and internal elastic net classification.

[0314] As used herein, a disease risk score is an indicator of the likelihood that a patient will be infected with a bacteria when the patient's composite expression score is compared to the composite expression score of a comparator group.

[0315] Develop disease risk scores into simple clinical tests for disease severity or disease risk prediction

[0316] The methods outlined above, in which a complex RNA expression signature of a disease or disease process is converted into a single score that is predictive of disease risk, can be used to develop simple, inexpensive, and clinically applicable tests for disease diagnosis or risk prediction.

[0317] The procedure is as follows: For tests based on differential gene expression between cases and controls (or different categories of cases, such as severity), the up- and down-regulated transcripts identified as relevant can be printed on a suitable solid surface, such as a microarray slide, microbeads, test tubes or wells.

[0318] Upregulated transcripts can be colocalized with downregulated transcripts in separate wells or separate tubes. A panel of unchanged housekeeping genes can also be printed separately to normalize the results.

[0319] RNA recovered from individual patients using standard recovery and quantification methods (with or without amplification) was hybridized to pools of up- and down-regulated transcripts as well as unchanged housekeeping transcripts.

[0320] Control RNA was hybridized in parallel to the same pool of up- or down-regulated transcripts.

[0321] A total value, e.g., fluorescence, is then read for the subject sample, and optionally the up- and down-regulated transcripts for the control sample, and the results are combined to give a composite expression score for the patient and control, which is then compared to a reference range for an appropriate number of healthy controls or comparison subjects.

[0322] Correct the detected signal for the relative abundance of RNA species in the subject sample

[0323] The details above explain how to reduce the complex features of many transcripts to a minimum set that can distinguish patients from other phenotypes to the greatest extent. For example, in the set of upregulated transcripts, there are some transcripts whose total expression levels are many times lower than the total expression levels of other transcripts. However, despite the overall low expression levels, these transcripts may have high discrimination. The weights derived from the net elasticity coefficient can be included in the test in a variety of different ways. First, the copy number of a single transcript included in the assay can vary. Secondly, in order to ensure that the signal from a rare important transcript is not overwhelmed by the signal from a transcript expressed at a higher level, one option is to select a probe that is neither overexpressed nor weakly expressed for testing to maximize the effect of multiple probes. Alternatively, the signal from a low abundance transcript can be adjusted by a scaling factor.

[0324] While this can be done at the analysis stage with current transcriptomics technologies, as each signal is measured individually, in simple colorimetric tests, only the total color change is measured, and therefore the signal cannot be scaled from selected transcripts. This problem can be addressed by reversing the chemistry typically associated with arrays. In conventional array chemistry, probes are coupled to a solid surface and the amount of biotinylated, patient-derived bound target is measured. Instead, we propose coupling biotinylated cRNA from the patient to an avidin-coated surface, and then adding DNA probes coupled to a chromogenic enzyme via an adapter system. During the design and manufacturing phase, probes targeting low-abundance but important transcripts are coupled to a greater amount or more potent form of the chromogenic enzyme, allowing the signal from these transcripts to be “amplified” in the final single-channel colorimetric readout. This approach will be used to normalize the relative input of each probe in the up-, down-, and housekeeping channels of the kit so that each probe makes an appropriately weighted contribution to the final readout, which can take into account its discriminatory power, as implied by the weights of the variable selection method.

[0325] The detection system for measuring multiple up-regulated or down-regulated genes can also be adapted to detect transcripts comprising diagnostic features using rTPCR (wherein the individual combined values ​​of up-regulated and down-regulated transcripts are summed) or physical detection methods (such as electrical impedance changes). In this method, the transcripts in question are printed on the nanowire surface or in a microfluidic cartridge, and the binding of the corresponding ligand for each transcript is detected by changes in impedance or other physical detection systems.

[0326] In one embodiment, the gene chip is a fluorescent gene chip, that is, the readout is fluorescence.

[0327] As used herein, fluorescence refers to the emission of light by a substance that has absorbed light or other electromagnetic radiation.

[0328] Therefore, in an alternative embodiment, the gene chip is a colorimetric gene chip, such as a colorimetric gene chip using microarray technology, wherein an enzyme (such as peroxidase or other chromogenic substrate) is attached to a biotin probe currently used to attach a fluorescent marker to DNA using avidin. The present disclosure extends to microarray chips that are suitable for reading by colorimetric analysis and are suitable for distinguishing subjects with bacterial infections from subjects with viral infections or inflammatory diseases. The present disclosure also extends to the use of colorimetric chips for analyzing subject samples to distinguish subjects with bacterial infections from subjects with viral infections or inflammatory diseases.

[0329] As used herein, colorimetry refers to an assay in which the output is in the human visible light spectrum.

[0330] In alternative embodiments, sets of genes or probes used to distinguish subjects with bacterial infections from subjects with viral infections or inflammatory diseases may be detected by physical detection methods including nanowire technology, electrical impedance changes, or microfluidics.

[0331] The reading of this assay can be converted from the fluorescence reading used in current microarray technology to a simple colorimetric format, or converted using a physical detection method (such as impedance change), which can be read with minimal equipment. For example, this can be achieved by utilizing biotin currently used to attach fluorescent markers to DNA. Biotin has a very high affinity for avidin, which can be used to connect enzymes such as peroxidase or other chromogenic substrates. This process will allow the use of a chromogenic process rather than fluorescence to quantify the number of cRNAs bound to the target transcript. Then, a simplified detection method providing a yes / no disease state indication can be developed by comparing the color intensity of the up-regulated and down-regulated libraries of the transcript with a control color standard. Similar methods can use physical methods (such as electrical impedance changes) to detect multiple gene features.

[0332] This aspect of the invention may be particularly advantageous for use in remote or resource-poor settings or for rapid diagnosis in "near patient" testing, for example in parts of Africa, because the equipment required to read the chip may be simpler.

[0333] As used herein, a multiplex assay refers to an assay that measures several analytes (usually dozens or more) simultaneously in a single run / cycle of the assay. Such an assay is distinguished from procedures that measure one analyte at a time.

[0334] In one embodiment, a custom gene chip for use in the method is provided, in particular, as described herein.

[0335] In one embodiment, a known gene chip is provided for use in the method described herein, specifically, for identifying one or more gene features described herein.

[0336] In one aspect, a method is provided for determining whether to administer a KD treatment to a subject (e.g., a subject suspected of having KD, e.g., a subject exhibiting symptoms of having KD) by using the methods disclosed herein, and administering the treatment to the subject when the methods indicate that the subject has KD. Examples of suitable treatments for KD include, but are not limited to, gamma globulin (IVIg), aspirin or other anti-inflammatory agents, such as steroids and infliximab, including combinations thereof.

[0337] Gene signature, gene transcript signature, gene panel, disease signature, diagnostic signature, and gene profile are used interchangeably throughout and should be interpreted to mean gene signature.

[0338] In the context of this specification, "comprising" should be interpreted as "including".

[0339] Aspects of the invention comprising certain elements are also intended to extend to alternative embodiments wherein the relevant elements "consisting of" or "consisting essentially of."

[0340] The embodiments of the present invention may be combined where technically appropriate.

[0341] Embodiments are described herein as comprising certain features / elements. The disclosure also extends to embodiments consisting solely of or consisting essentially of said features / elements.

[0342] Technical references such as patents and applications are incorporated herein by reference.

[0343] Any embodiment specifically and explicitly described herein may form the basis of a disclaimer either alone or in combination with one or more further embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0344] Figure 1 The diagnostic algorithm used to assign patients to diagnostic groups is shown. KD = Kawasaki disease; AHA = American Heart Association; CAA = Coronary Artery Aneurysm; JIA = Juvenile Idiopathic Arthritis; HSP = Henoch-Schönlein Purpura; CRP = C-Reactive Protein.

[0345] Figure 2 The overall study workflow is shown, which illustrates sample processing, derivation of test and training data sets, data processing and analysis procedures.

[0346] a See Methods (Example 1); b Healthy controls were used for model construction but were not included in the assessment of model accuracy; c The diagnostic performance was evaluated in 72 patients (disease days 2-7). Abbreviations: KD = Kawasaki disease; DB = confirmed bacterial; DV = confirmed viral; U = infection of uncertain bacterial or viral etiology; JIA = juvenile idiopathic arthritis; HSP = Henoch-Schonlein purpura; HC = healthy control; PDMS = parallel deterministic model search; SDE = significantly differentially expressed; FC = fold change.

[0347] Fig. 3 shows the performance of 13 transcript features on the discovery test set and validation set. The classification (A) and receiver operating characteristic (ROC) curve (B) of 13 transcript features in the discovery test set (including patients with Kawasaki disease (KD) and patients with other diseases) are used. Disease risk score (DRS) is used. The classification (C) and ROC curve (D) of 13 transcript features in the validation set (including three KD clinical subgroups with different diagnostic certainties and patients with other diseases). In the box plot, the horizontal line represents the median; the lower edge and the upper edge represent the interquartile range; the whisker represents the range, or 1.5 times the interquartile range, whichever is smaller. The horizontal line on the figure represents the DRS threshold, which is predicted to be separated by patients with KD (above the straight line) or patients without KD (below), as determined by the points in the ROC curve, which maximizes the sensitivity and specificity in the discovery training group. KD = Kawasaki disease; DB = confirmed bacterial; DV = confirmed viral; U = infection of uncertain bacterial or viral etiology; JIA = juvenile idiopathic arthritis; HSP = Henoch-Schönlein purpura; KD = definitive KD; KD-HP = highly probable KD; KD-P = possible KD.

[0348] Figure 4 The performance of 13 transcripts divided by days of illness at sample collection in the validation set is shown. The X-axis shows the day of sample collection relative to the first day of illness (i.e., onset of fever). Black dots = confirmed KD, gray dots = highly probable KD, black dots with arrows = possible KD clinical subgroups in the validation set.

[0349] Figure 5 The principal component analysis (PCA) plots of PC1 and PC2 in the discovery cohort after background adjustment and normalization are shown. Samples from KD patients were removed from subsequent analysis (arrows). Each point is data from the array. KD = Kawasaki disease, DB = confirmed bacterial, DV = confirmed viral, HC = healthy control, U = infection of uncertain bacterial or viral etiology, JIA = juvenile idiopathic arthritis, HSP = Henoch-Schönlein purpura.

[0350] Figure 6 PCA plots of (A) initial merge of validation cohorts and (B) merged using ComBat are shown. Each point represents data from the array; KD-acute = acute Kawasaki disease, KD-conv = convalescent Kawasaki disease, DB = confirmed bacteria, DV = confirmed virus, U = infection of uncertain etiology of bacteria or virus, HC healthy controls. Image (B) contains data from samples of 30 KD patients after day 7 after fever, which were not included in the diagnostic performance calculations.

[0351] Figure 7A gene network from a 13 transcript feature is shown. The network was generated using Ingenuity PathwaysAnalysis. 12 of the 13 transcripts were mapped to the database. The network contains 7 focal molecules, which are the most important networks in the analysis. Each molecule is colored according to the direction of expression in the KD. Solid lines represent direct interactions and dashed lines represent indirect interactions. The legend for this network is available at: http: / / ingenuity.force.com / ipa / articles / Feature_Description / Legend.

[0352] Example 1-13 Identification of transcript gene features

[0353] Patient study group

[0354] The differential diagnosis of KD includes a variety of infectious and inflammatory conditions, so we established a case-control cohort to study children with KD and a range of other infectious and inflammatory diseases with clinical symptoms overlapping with KD. If patients had a febrile illness and required blood testing for clinical investigations, they were pre-enrolled at pediatric centers in the United Kingdom, the Netherlands, Spain, and the United States as part of the Immunopathology of Respiratory, Inflammatory and Infectious Disease Study

[20] , the GENDRES study in Spain, the Kawasaki Disease Research Center Program in the United States, or the Kawasaki Study in the Netherlands.

[0355] The children with KD recruited represented children presenting directly to the emergency department of the study center, as well as patients referred from regional centers. However, our study included only patients who had blood samples collected before starting KD treatment with IVIG and during the first 7 days of illness. Febrile controls were recruited with blood samples collected early, before clinical diagnosis was confirmed, to obtain samples as close to presentation as possible, and included patients referred by practitioners in the community for evaluation of possible KD, who represent a real population for whom diagnostic testing would be of great relevance.

[0356] Once the results of all investigations were available, febrile controls were assigned to diagnostic groups according to predefined criteria (Supplementary Appendix and Figure 1 Children with comorbidities that could affect gene expression, such as immunosuppressive therapy or bone marrow transplantation, were excluded. Comparison groups of children with inflammatory diseases were included: Henoch-Schönlein purpura (HSP) and juvenile idiopathic arthritis (JIA).

[0357] Patients in the validation study group were recruited similarly as previously described as part of a biomarker study of febrile children who presented to the hospital and required blood testing [21, 22]. Patients who presented to the hospital within ten days of the onset of a febrile illness were recruited, and blood samples were collected for gene expression analysis at the same time as routine diagnostic studies to assess the etiology of the children. Healthy control children with no recent (2 weeks) history of fever or immunization, as well as KD and febrile control patients, were recruited as part of the discovery and validation studies. Data from healthy controls were used to normalize the data obtained in the different microarray experiments but were not used to evaluate the performance of the signature.

[0358] KD Case Definition

[0359] KD was diagnosed according to the American Heart Association (AHA) criteria

[14] . Patients diagnosed with KD underwent 2D echocardiography at presentation and at two and six weeks after onset. Patients were included as having incomplete KD if the maximum coronary artery Z score (Zmax) (standard deviation units of the mean inner diameter relative to the mean body surface area) of the left anterior descending or right coronary artery was ≥2·5 at any time during the disease course, or if patients with fewer than four of the classic criteria fulfilled the algorithm for incomplete KD in the AHA guidelines. Patients were classified as having normal coronary arteries (Zmax <2·5) or dilated coronary arteries (Zmax ≥2·5 <5·0) or CAA (Zmax ≥5·0). Because of inter-operator variability in the exact coronary artery size, we set a higher threshold (Zmax ≥5.0) to define patients with aneurysms to reduce misclassification.

[0360] Further classification of KD by diagnostic certainty

[0361] Because there is no “gold” standard for diagnosing KD, some patients may meet the criteria for KD but have other conditions, such as staphylococcal or streptococcal infection, viral infection, or inflammatory disease. Therefore, we further divided patients with KD into validation study groups based on the certainty of the clinical diagnosis. All clinical records, laboratory results, echocardiogram reports, response to treatment, and follow-up clinical notes were reviewed by an independent pediatric infectious disease specialist and KD expert (author MPG–blind to the analyses). Patients with documented CAA (Zmax ≥ 5·0) for six weeks after onset were considered to have definite KD, as there was no other self-resolving inflammatory disease that could cause CAA in childhood. The remaining patients (all of whom were treated with IVIG for suspected KD by the clinical team) were classified by the expert reviewers as very likely, possible, or unlikely to have KD. No cases of “unlikely KD” were identified by this review.

[0362] Febrile control children with infection or other inflammatory syndromes

[0363] use Figure 1 Children with a febrile illness were identified as having a confirmed bacterial infection, a confirmed viral infection, a suspected bacterial or viral infection, HSP, or JIA according to the criteria shown and described in the Supplementary Appendix.

[0364] Ethical approval and consent

[0365] Patients were recruited with approval from the UCSD Research Ethics Committee (Human Research Protection Program #140220), Spain (Galicia Clinical Research Ethics Committee, CEIC reference 2010 / 015), Amsterdam (NL41023.018.12 and NL34230.018.10), and the United Kingdom (St. Mary's Hospital 09 / H0712 / 58, 13 / LO / 0026).

[0366] Overseeing and conducting research

[0367] All outcomes were assessed by at least two independent clinicians (authors JAH, JCB, JK, MPG, AMB) who were not involved in the patients' care and who were then classified into disease groups ( Figure 1 ). All samples were anonymized. Transcriptomic data sets were analyzed only after clinical assignment was completed and independent verification was sent (Supplementary Appendix).

[0368] Discovery and validation of gene expression signatures

[0369] The overall study design and feature discovery process are as follows Figure 2 Whole blood was collected into PAXgene Blood RNA tubes (PreAnalytiX, Germany) at recruitment (before IVIG treatment for KD cases), frozen, extracted, and analyzed on human HT-12v.4 BeadChip arrays (Illumina). In a subset of the validation study group, an earlier Illumina BeadChip array (HT-12v.3) with a large number of overlapping probes was used. Details of the laboratory methods are provided in the Supplementary Appendix.

[0370] Statistical analysis

[0371] Discovery of transcript signatures

[0372] Transcriptomics data were analyzed using the “R” language and environment for statistical computing (R) 3.2.2 (Supplementary Methods). Figure 2As shown, the discovery study set was randomly divided into an 80% “training” set and a 20% “testing set”. Features were identified in the training set and validated in the test set as well as in a second study set (validation study set) established using previously reported acute and convalescent KD patients

[21] and acute bacterial and viral patients

[22] (Supplementary Methods). After quality control and filtering (Supplementary Methods), transcripts that were significantly differentially expressed (SDE) in KD patients compared to all other diseases were identified in the training set.

[0373] Small feature discovery using Parallel Deterministic Model Search (PDMS)

[0374] A novel method, PDMS, which identifies and ranks transcript signatures based on the fewest number of transcripts and highest discrimination accuracy, was used to identify a parsimonious gene expression signature comprising the minimum number of transcripts that best distinguishes KD from other diseases. The method first evaluates all possible single-gene and two-gene models that distinguish KD from control diseases based on all SDE transcripts, and when an additional gene is added to the model, the 100 most suitable two-gene models are carried forward to the next round and all combinations are evaluated again. This process proceeds as follows: additional genes are gradually added to the best 100 models one at a time. The best signature for a given number of transcripts (model size) is selected after ranking each model by the Watanabe-Akaike Information Criterion, which is a Bayesian estimate of the out-of-sample error

[23] . The optimal model size is determined by cross-validation. Further details are in the Supplementary Statistical Methods.

[0375] Disease risk scoring and model accuracy assessment

[0376] We applied a previously reported disease risk score (DRS) approach that assigns individual disease risk based on the transcripts included in the diagnostic signature

[15] . DRS combines the fluorescence intensity of upregulated transcripts and subtracts the combined fluorescence intensity of downregulated transcripts

[15] , which may be helpful for test development of complex signatures. Healthy controls were used for model construction but were not included in the assessment of model accuracy, which was instead assessed by the area under the receiver operator curve (AUC), sensitivity, and specificity.

[0377] Supplementary methods

[0378] RNA sample extraction and processing

[0379] Whole blood (2.5 ml) was collected into PAXgene Blood RNA tubes (PreAnalytiX, Germany), incubated for 2 hours, frozen at -20°C within 6 hours of collection, and then stored at -80°C. RNA was extracted using the PAXgene Blood RNA Kit (PreAnalytiX, Germany) according to the manufacturer's instructions. The integrity and yield of total RNA were assessed using an Agilent 2100 Bioanalyzer and a NanoDrop1000 spectrophotometer. The samples used in the discovery cohort were from the United States (UCSD), Spain, the Netherlands, and the United Kingdom. All samples, except those from the United States, were extracted in the United Kingdom. After quantification and quality control, biotin-labeled cRNA was prepared from 500 ng of RNA using the Illumina TotalPrep RNA Amplification Kit (Applied Biosystems). The labeled cRNA was hybridized overnight to the Human HT-12v.4 Expression BeadChip Array (Illumina). After washing, blocking, and staining, the array was scanned using the Illumina BeadArray Reader according to the manufacturer's instructions. Microarray images were checked for artifacts and QC parameters were assessed using Genome Studio software. No arrays were excluded at this stage.

[0380] Pathogen diagnosis

[0381] Viral diagnostics were performed on nasopharyngeal aspirates using immunofluorescence (RSV, adenovirus, parainfluenza, influenza A+B) and nested PCR (RSV, adenovirus, parainfluenza 1-4, influenza A+B, bocavirus, metapneumovirus, rhinovirus / enterovirus). Bacterial cultures included blood, CSF, urine, and tissue sites. Pneumococcal antigens were measured in blood and urine, and bacterial DNA was detected by meningococcal and pneumococcal PCR.

[0382] Fever control diagnostic process

[0383] Patients underwent diagnostic workup as directed by the clinical team, including blood count, blood chemistry, C-reactive protein (CRP), urine, and throat swab cultures; cerebrospinal fluid analysis and chest radiographs were performed where appropriate. Multiplex PCR was used to detect common respiratory viruses in nasopharyngeal aspirates or throat swabs and common viruses in the blood. Once the results of all investigations were available, the following predetermined criteria were followed: Figure 1 ) to assign patients to diagnostic groups.

[0384] Bacterial Infections:

[0385] Patients assigned to the bacterial pathogen group had bacterial pathogens (Gram-positive cocci or Gram-negative bacilli) identified by culture or by molecular techniques in samples from sterile sites (blood, CSF, pleural cavity, joints, urine), and a clinical syndrome consistent with the identified bacterial species. This group included patients with and without viral co-infections. Children diagnosed with other bacterial infections (e.g., mycoplasma, pertussis, mycobacterium) were not included in this group. No threshold for inflammatory markers was set for this group, as identification of bacteria in samples from sterile sites was used as conclusive evidence for a confirmed bacterial infection.

[0386] Viral Infections:

[0387] Patients in the viral infection group had an identified virus, a clinical syndrome consistent with viral infection, and no microbiological or clinical features of bacterial disease. To avoid including children with occult bacterial infection in the viral group, children with elevated inflammatory markers were excluded. Maximum thresholds were set at 60 mg / L for CRP and 12 x 10 for neutrophil count. 9 Among the 94 children, the most common pathogens were RSV (27 children), influenza A / B, and adenovirus (23 children each).

[0388] Undetermined bacterial or viral infection:

[0389] Children with acute febrile illness and features of infection were labeled “uncertain bacterial or viral” if they could not be confidently assigned to one of the above groups. Children in this group had features of uncertain bacterial or viral infection, negative microbiologic findings or lack of virologic testing, syndromes inconsistent with microbiologic findings, inflammatory markers inconsistent with other clinical features of illness, or insufficient clinical data to be confidently coded in another group. Patients in this group had no bacterial infection identified at a sterile site, and some did have detectable viruses.

[0390] Other inflammatory syndromes:

[0391] a) Henoch-Schönlein purpura (HSP) was diagnosed in children with frank purpura, usually on the buttocks and extensor surfaces, accompanied by abdominal pain, arthralgia, or renal abnormalities (hematuria and proteinuria); b) Juvenile idiopathic arthritis (JIA) was defined according to the International Federation of Rheumatology

[37] . Patients with JIA were included: i) treatment naïve; and ii) actively exacerbating / stagnation.

[0392] Statistical methods

[0393] Microarray preprocessing - discovery dataset

[0394] Background subtraction and robust spline normalization (RSN) were applied to the raw expression data using the R package lumi

[38] . Sample outliers were assessed by principal component analysis (PCA). One sample from a Kawasaki patient was clearly an outlier on PC 1 and was removed from the analysis ( Figure 5 ).

[0395] The samples in the discovery dataset were randomly assigned to ten different folds, with the condition that the number of each comparison group was equal (KD Kawasaki disease, DB confirmed bacteria, DV confirmed virus, U bacterial or viral infection of uncertain etiology, JIA juvenile idiopathic arthritis, HSP Henoch-Schönlein purpura, HC healthy controls). Two folds (20%) were retained as test sets, while the remaining eight folds formed the training set. Because diagnostic testing for KD is most important early in the disease, we developed features using only samples from patients with fever for 7 days or less in the discovery cohort.

[0396] Microarray preprocessing - validation dataset

[0397] The validation dataset was constructed by merging two gene expression datasets: one with acute and convalescent Kawasaki samples

[39] and the other with bacterial and viral infections

[40] . All convalescent samples had ESR (erythrocyte sedimentation rate) levels below 40 mm / h and all acute samples were collected within ten days of onset. Background subtraction and RSN normalization were applied to both datasets separately in the R package lumi

[38] . At this stage, there were differences between the different cohorts. This is evident from the PCA plots, which show that PC1 clearly differentiates samples by batch ( Figure 6 a). Therefore, we used the ComBat

[41] method to eliminate the batch effect. Two binary covariates were passed to ComBat, which classified the samples into three groups - healthy, KD, and other diseases. Kawasaki convalescent samples were classified as healthy. PCA after ComBat showed that the samples from the two batches overlapped in the plot of PC1 versus PC2, with no obvious batch effect ( Figure 6 b).

[0398] Model estimation

[0399] Prior to model estimation, probes were pre-filtered to identify robustly expressed transcripts with a log2 fold change ≥ 1 between relevant disease groups. This was implemented by selecting probes in the training data that met all of the following criteria:

[0400] 1. Probes measured on both V3 and V4 Illumina Beadchips

[0401] 2. Robustly expressed transcripts: For each probe, we calculated the proportion of samples with a detection threshold p-value < 0.01 in each comparison group and selected probes with this proportion greater than 80% in at least one disease group.

[0402] 3. Most Kawasaki patients were recruited from UCSD. To ensure that probe selection was not biased by batch effects arising from UCSD, we excluded probes associated with UCSD recruitment at p < 0.05 in the linear model conditioned on age in months, as well as all disease groups (DV, U, KD, and HSP) that also included non-KD patients recruited from UCSD.

[0403] 4. Calculate log2 fold changes between Kawasaki and other comparison groups (depending on age); we propose probes with log2 fold change |>1 for at least one of these comparisons

[0404] The probe association statistics used in steps (3) and (4) above were calculated using the functions lmFit and eBayes in the R package limma

[42]

[0405] Discovery using Parallel Deterministic Model Search (PDMS)

[0406] We used an in-house method, PDMS, to obtain a parsimonious gene expression signature that balances a small number of transcripts with accurate discrimination. The method iteratively estimates logistic regression coefficients for a selected subset of gene expression levels (covariates). The regression coefficients are assigned a zero-centered Gaussian prior distribution with a precision parameter of τ (where τ = 1 / variance and equal to the penalty) to shrink the coefficients to zero. The method searches for as many models as possible and selects the "best" model, each of which includes a unique subset of the selected covariates and their respective logistic regression coefficients.

[0407] To find the best shrinkage parameter for the prior probability distribution, we evaluated the accuracy of each model using LASSO cross-validation on multiple partitions of the discovery data. The R package glmnet was used to determine the LASSO penalty with the minimum out-of-sample cross-validation bias. We then selected the model with the largest regression coefficient (β) from the best LASSO fit. Max ) to set τ by equating the penalty induced by the LASSO with the penalty induced by the Gaussian prior. Let λ represent the LASSO penalty parameter:

[0408] LASSO penalty = Gaussian penalty

[0409]

[0410] The PDMS method operates as follows: using pre-filtered probes that are robustly expressed between groups (log2 fold change ≥ 1), all possible single-transcript models and two-transcript models are evaluated and ranked according to the log-likelihood (a measure of how well the model fits the data), and the top 100 two-transcript models are proposed. In the next stage, the algorithm determines a unique set of three-gene models that can be constructed by adding one gene to each of the top 100 two-gene models. The log-likelihood of these models is calculated, and then the construction of the top 100 models is continued to construct a model with a gene larger than one.

[0411] For models of a given size (number of transcripts), the models were ranked by the Watanabe-Akaike Information Criterion (WAIC)

[43] . WAIC is a Bayesian information criterion for estimating the expected out-of-sample error that penalizes models based on the number of effective parameters and assumes inference is made from the posterior distribution. PDMS makes inferences based on posterior means, so WAIC is appropriate for our application. For details on how WAIC is estimated, see Gelman et al.

[44] . Although WAIC adjusts for overfitting by adding a correction for the number of effective parameters, it does not address the increased false positive rate that results from searching many models. Across all model sizes explored, PDMS does not simply select the model with the lowest WAIC, but instead introduces an additional penalty that minimizes:

[0412] Corrected standard = WAIC + αk

[0413] Where k is the model size and we take α=1.

[0414] Calculating model accuracy

[0415] The area under the ROC curve and the corresponding confidence intervals for the model applied to the test and validation datasets were calculated using the R package pROC

[45] .

[0416] The results for each patient were summarized into a disease risk score (DRS) to determine the accuracy of classification using the 13 transcript features, and the optimal threshold for classification as KD or non-KD based on the training data was determined by the point in the ROC curve that maximized the distance to the identity line (maximum value of (sensitivity + specificity)) according to Youden's J statistic

[46] . The same threshold was used in the accuracy calculation of the validation data.

[0417] Confidence intervals (CIs) for sensitivity and specificity were calculated using the Jeffrey method. The Jeffrey method is derived from a Bayesian perspective in which the basis of interest is assigned to a non-informative Jeffrey reference prior - β(1 / 2, 1 / 2).

[47] Thus, the 95% CI for sensitivity is derived from the 2.5% and 97.5% quantiles of the β(p+1 / 2, q+1 / 2) distribution, where p is the number of true positives and q is the number of false negatives.

[0418] result

[0419] The number of patients in each diagnostic category is as follows Figure 2 The clinical and demographic characteristics of KD patients are shown in Table 1 , and the clinical characteristics and demographics of patients with other inflammatory syndromes and infections are shown in Tables 3-6 . Principal component analysis of normalized gene expression profiles was performed on the discovery group (training group and test group) and validation group, respectively; Figure 5 and 6 PC1 versus PC2 for both analyses are plotted. After combining KD with case-control data using the ComBat algorithm, the study groups were clustered in the discovery and validation groups

[24] (see Supplementary Statistical Methods above). Table 1: Clinical characteristics and laboratory values ​​at acute time points for KD subjects in the discovery and validation study groups

[0420]

[0421]

[0422] All values ​​are shown as median (IQR). There were no significant differences between the characteristics of the discovery and validation patients. a: sickness day 1 = first day of fever; b: hemoglobin normalized by age; c: >140 is written as 140; d: among 102 KD patients, 30 patients with sickness days ≥8 days at sampling were excluded, and the remaining 72 patients were used for diagnostic performance; e P value for discovery vs validation = 0.051.

[0423] Identifying minimal transcript features

[0424] There were 1600 transcripts that passed QC that were significantly differentially expressed between KD and all other diseases and healthy controls (defined as |log2 fold change|>1 in KD vs. at least one comparison group). To determine the minimum features suitable for testing, PDMS was next used for variable selection. This approach identified 13 transcript features (Table 2), which, when implemented as DRS, had the following diagnostic performance: AUC in the test set was 96.2% (95% CI, 92.5%, 99.9%), and sensitivity / specificity were 81.7% (95% CI, 60.0%, 94.8%) and 92.1% (95% CI, 84.0%, 97.0%), respectively ( Figure 3A , B).

[0425] Table 2: Genes included in the diagnostic signature

[0426]

[0427] The logistic regression coefficient represents the ability of a gene to discriminate KD in the PDMS model; genes with positive values ​​have increased expression in KD relative to other diseases; genes with negative values ​​have decreased expression in KD.

[0428] Feature performance in the validation set

[0429] When the signature was applied to 72 KD cases in the validation set, the AUC was 96.5% (95% Ci, 93.7%, 99.3%), the sensitivity was 90.8% (95% Ci, 82.5%, 96.2%), and the specificity was 89.1% (95% CI, 83.0%, 93.7%). Since the clinical features of KD overlap with other conditions, and since any KD study group may contain patients without KD, we assessed whether the certainty of the clinical diagnosis corresponds to the strength of the KD DRS prediction score. The performance of the 13 transcript signatures in the confirmed, highly probable, or possible KD patients in the validation set (see Methods) followed the clinical certainty of the diagnosis. When analyzed separately, the performance of the 13 transcript PDMS signature followed the clinical certainty of diagnosis in the definite, highly likely, or possible KD groups, with ROC AUCs of 98.1% (95% CI, 94.5%, 100%), 96.3% (95% CI, 93.3%, 99.4%), and 70.0% (95% CI, 53.4%, 86.6%), respectively. Figure 3C , D).

[0430] Characteristic performance by sick days

[0431] The discovery set contains KD patients who were sick for up to 7 days (day 1 is the first day of fever), and the feature has been validated on patients who were sick for up to 7 days (including day 7). When the feature was applied to 30 patients on days 8-10 of illness, the performance of the feature decreased ( Figure 4 ).

[0432] discuss

[0433] We have identified a 13-transcript signature that distinguishes KD from patients with bacterial, viral, and inflammatory diseases. The signature has high sensitivity and specificity for the early diagnosis of KD, suggesting that it may form the basis of a diagnostic test. Our findings extend previous gene expression studies in KD that have focused on immune pathogenesis [21, 25-29]. For 5 of the 13 transcripts in the signature, expression was lower in KD patients than in the non-KD group (Table 2). Of these 5 transcripts, S100 calcium-binding protein P (S100P) has previously been reported to increase expression in KD in the acute phase compared to the recovery phase

[30] or viral infection [29, 30]. S100P expression was highest in bacterial patients, and the selection of this transcript in the PDMS model was driven by KD bacterial discrimination. An interferon-induced gene, interferon α-induced protein 27 (IFI27), which regulates apoptosis, has been reported to be upregulated in febrile children with viral infections compared with children with acute bacterial infections

[31] and autoimmune diseases [32,33]. Low transcript abundance of gene families induced by type 1 interferons has been previously reported in whole blood gene expression comparisons in acute KD versus adenovirus infection

[29] , consistent with the inclusion of IFI27 in the model as a negative predictor of KD. CD163 is a transmembrane receptor expressed in macrophages and monocytes involved in bacterial clearance during the acute phase of infection

[34] . Network analysis of the signature using Ingenuity Pathways Analysis showed that 7 of the 13 transcripts in the signature were connected together in a network centered around the TNF and IL6 hubs ( Figure 7 ).

[0434] Currently, the diagnosis of KD relies on four of five characteristic clinical criteria. If coronary artery abnormalities (dilatation or aneurysm) are detected on echocardiography, the fewer criteria are accepted as a diagnosis. Children with “incomplete KD” who do not meet the classic diagnostic criteria but who have prolonged fever and inflammation are at increased risk for CAA

[35] . One reason for the higher risk of CAA in incomplete KD is that the diagnosis is often delayed in patients who lack all clinical features. Because the clinical features of KD overlap with many other common childhood conditions (eg, staphylococcal and streptococcal toxin illnesses, viral fevers, Stevens Johnson syndrome, systemic juvenile idiopathic arthritis, and drug reactions

[36] ), treatment with IVIG may be delayed while waiting to exclude other conditions. Conversely, because the diagnosis of KD is considered in the differential of many childhood febrile illnesses and the consequences of delayed treatment can be serious, overtreatment with IVIG or second-line treatment with immunosuppressants may occur. A diagnostic test that accurately distinguishes KD from other infectious and inflammatory processes would be a major advance in the management of the disorder, reducing unnecessary testing and inappropriate treatment and allowing earlier treatment with IVIG and other anti-inflammatory agents.

[0435] In establishing the discovery and validation study groups, we aimed to include a variety of disorders that have overlapping features with KD, including infectious and inflammatory diseases. The features we have identified distinguish KD from a variety of other conditions. Because KD is diagnosed based on a range of clinical features and no gold standard for diagnosis exists, it is difficult to evaluate biomarkers or tests. It is likely that some patients with non-KD diseases with overlapping clinical features will be included in the cohort of all children treated with IVIG for presumed KD. To assess the correspondence of the KD DRS with the level of diagnostic certainty, we classified all patients in the validation set as definite, highly probable, or possible KD based on independent review of all clinical data. The features we observed in the definite and highly probable groups had higher sensitivity and specificity than those observed in the possible group. The diagnostic accuracy of the KD-specific features is ready to be tested in prospective studies.

[0436] We recognize the strengths and limitations of this study. First, the epidemiology of KD varies globally by ethnicity, with higher prevalence in East Asia and lower prevalence in Europe. Further studies are needed to investigate whether there are ethnic and geographic differences in gene expression in KD. A strength of our study is that the signature was developed from a training set that includes a range of ethnicities. Febrile control samples in the discovery set were obtained from all centers, while multiethnic KD samples were obtained from UCSD. Second, in the validation experiments, KD and case-control data from different Illumina microarray versions were combined by applying the ComBat algorithm

[24] and normalized to healthy control data for each platform. This normalization can reduce experimental and biological sources of variation between datasets, and therefore, the accuracy of the diagnostic signature when applied to the validation set (AUC results) may be underestimated compared to the validation dataset obtained from a single microarray experiment. Third, 13 transcript signatures were discovered using KD patients with illness for no more than 7 days. A strength of our signature is that it may help to make early KD diagnoses before 5 days of fever onset. However, further work is needed to identify the best signature to diagnose late “missed” KD patients.

[0437] Translating a multi-transcript signature into a rapid clinical test for use in hospital diagnostic laboratories is challenging but is more achievable due to the relatively small number of transcripts in our signature and the rapidly evolving technology for detecting nucleic acids. In addition, DRS provides a new approach to individual disease risk assignment without the need for complex analyses and provides a platform for the development of tests that include up- or down-regulated transcripts of the KD signature in the same location and detect a combined signal of both.

[0438] "Our study shows that KD can be distinguished from a range of infectious and inflammatory conditions, which are often confused clinically using small numbers of transcripts in the blood. The development of a rapid test based on this gene expression signature would be a major advance allowing earlier treatment and therefore prevention of cardiac complications of this serious childhood disease. Our findings represent an advance towards better diagnosis of the disease based on molecular signatures rather than clinical criteria and are therefore relevant to many other clinical syndromes."

[0439] Data Repository

[0440] The data discussed above have been deposited in the Gene Expression Omnibus at NCBI (Edgar et al., 2002) and are accessible through GEO Series accession number GSE73464 ( http: / / www.ncbi.nlm.nih.gov / geo / )access.

[0441] Supplementary Table

[0442] Table 3: Clinical characteristics of children in the juvenile idiopathic arthritis cohort (Findings)

[0443]

[0444] a All values ​​are expressed as median (IQR); b Laboratory values ​​for 27 patients in the initial treatment set and 35 patients in the active exacerbation / stagnation set. ESR = erythrocyte sedimentation rate, ANA = antinuclear antibodies, ANCA = antineutrophil cytoplasmic antibodies.

[0445] Table 4: Clinical characteristics (findings) of children in the Henoch-Schonlein purpura group

[0446]

[0447] a All values ​​are expressed as median (IQR); b Hemoglobin was normalized for age; c Laboratory data were available from 15 patients; d Laboratory data were available from 14 patients; e Laboratory data were available from 4 HSPs; f Laboratory data were available for 8 patients; ESR = erythrocyte sedimentation rate

[0448] Table 5: Clinical characteristics of children with bacterial and viral infections, infections of uncertain bacterial or viral etiology, and healthy controls (discovery and validation)

[0449]

[0450] a All values ​​are shown as median (IQR); b Percentage of patients whose race was known, c Until the study blood sampling, d The highest values ​​of CRP in the disease were reported.

[0451] Table 6: Viral and bacterial pathogens of patients in the confirmed bacterial and viral groups

[0452]

[0453] Table 7: Summary of model performance

[0454]

[0455] Example 2 - Identification of gene features with few transcripts

[0456] Alternative smaller signatures (with fewer transcripts) were generated based on a subset of the original 13 transcripts using PReMS software (Hoggart, 2018).

[0457] PReMS searches for many logistic regression models built from the best subset of biomarkers, iteratively increasing the size of the model. A zero-centered Gaussian prior distribution is assigned to all regression coefficients to induce shrinkage. The method estimates the optimal shrinkage parameter, the best model for each model size, and the optimal model size.

[0458] Table 8 below shows examples of smaller 5, 6, 7, 8, 9, 10, 11 and 12 transcript features based on transcript combinations from the original 13 transcripts. The AUC values ​​for the test and validation data sets (see Example 1) are shown for each feature. Please note that for brevity, the sample list of gene features shown here is not exhaustive, but is exemplary only.

[0459] Table 8 - Examples of smaller features based on a subset of the original 13 transcripts

[0460]

[0461]

[0462]

[0463]

[0464]

[0465]

[0466]

[0467]

[0468]

[0469] Thus, this example demonstrates that smaller gene signatures based on subsets of 5, 6, 7, 8, 9, 10, 11, or 12 of the original 13 transcripts have good discriminatory power and can reliably identify individuals with KD from those without KD.

[0470] References

[0471] 1. Kawasaki T, Kosaki F, Okawa S, Shigematsu I, Yanagawa H. "A new infantile acute febrile mucocutaneous lymph node syndrome (MLNS) prevailing in Japan." Pediatrics. 1974; 54(3): 271-6. Epub 1974 / 09 / 01. PubMed PMID: 4153258.

[0472] 2. Makino N, Nakamura Y, Yashiro M, Ae R, Tsuboi S, Aoyama Y, et al. “Descriptive epidemiology of Kawasaki disease in Japan, 2011-2012: from the results of the 22nd nationwide survey.—” Journal of epidemiology / Japan Epidemiological Association. 2015;25(3):239-45. doi:10.2188 / jea.JE20140089. PubMed PMID: 25716368; PubMed Central PMCID: PMCPMC4341001.

[0473] 3. Du ZD, Zhao D, Du J, Zhang YL, Lin Y, Liu C, et al. “Epidemiologic study on Kawasaki disease in Beijing from 2000 through 2004.” The Pediatric Infectious Disease Journal. 2007;26(5):449-51.doi:10.1097 / 01.inf.0000261196.79223.18.PubMedPMID:17468660.

[0474] 4. Kim GB, Park S, Eun LY, Han JW, Lee SY, Yoon KL, et al. “Epidemiology and Clinical Features of Kawasaki Disease in South Korea, 2012-2014.” J Pediatr Infect Dis 2017;36(5):482-5. Epub 2016 / 12 / 21.doi:10.1097 / INF.0000000000001474. PubMed PMID:27997519.

[0475] 5. Hoggart CJ (2018). PReMS: “Parallel Regularised Regression Model Search for sparse bio-signature discovery.” bioRxiv 355479; doi: https: / / doi.org / 10.1101 / 355479.

[0476] 6. Harnden A, Mayon-White R, Perera R, Yeates D, Goldacre M, Burgner D. “Kawasaki disease in England: ethnicity, deprivation, and respiratory pathogens.” J Pediatr Infect Dis 2009;28(1):21-4. PubMed PMID:19145710.

[0477] 7. Holman RC, Belay ED, Christensen KY, Folkema AM, Steiner CA, Schonberger LB. “Hospitalizations for Kawasaki syndrome among children in the United States, 1997-2007.” J Pediatr Infect Dis. 2010;29(6):483-8. doi:10.1097 / INF.0b013e3181cf8705. PubMed PMID:20104198.

[0478] 8. Kato H, Sugimura T, Akagi T, Sato N, Hasino K, Maeno Y, et al. “Long-term consequences of Kawasaki disease.” “A 10-to 21-year follow-up study of 594 patients.” Circulation. 1996;94(6):1379-85. Epub 1996 / 09 / 15. PubMed PMID:8822996.

[0479] 9. Suda K, Iemura M, Nishiono H, Teramachi Y, Koteda Y, Kishimoto S, et al. “Long-term prognosis of patients with Kawasaki disease complicated by giant coronary aneurysms: a single-institution experience.” Circulation 2011;123(17):1836-42. doi:10.1161 / CIRCULATIONAHA.110.978213. PubMed PMID:21502578.

[0480] 10. Daniels LB, Gordon JB, Burns JC. “Kawasaki disease: late cardiovascular sequelae.” Current opinion in cardiology. 2012;27(6):572-7. doi:10.1097 / HCO.0b013e3283588f06. PubMed PMID:23075819.

[0481] 11. Yu JJ. “Use of corticosteroids during acute phase of Kawasaki disease.” World J Clin Pediatr. 2015;4(4):135-42. doi:10.5409 / wjcp.v4.i4.135. PubMed PMID: 26566486; PubMed Central PMCID: PMCPMC4637804.

[0482] 12. Tremoulet AH, Jain S, Jaggi P, Jimenez-Fernandez S, Pancheri JM, Sun X, et al. “Infliximab for intensification of primary therapy for Kawasaki disease: a phase 3 randomised, double-blind, placebo-controlled trial.” Lancet. 2014;383(9930):1731-8. doi:10.1016 / S0140-6736(13)62298-9. PubMed PMID:24572997.

[0483] 13. Dominguez SR, Anderson MS, El-Adawy M, Glode MP. “Preventing coronary artery abnormalities: a need for earlier diagnosis and treatment of Kawasaki disease.” Pediatr Infect Dis J. 2012;31(12):1217-20. Epub 2012 / 07 / 05.doi:10.1097 / INF.0b013e318266bcf9. PubMed PMID:22760536.

[0484] 14. McCrindle BW, Rowley AH, Newburger JW, Burns JC, Bolger AF, Gewitz M, et al. “Diagnosis, Treatment, and Long-Term Management of Kawasaki Disease: A Scientific Statement for Health Professionals From the American Heart Association.” Circulation 2017;135(17):e927-e99. Epub 2017 / 03 / 31.doi:10.1161 / CIR.0000000000000484. PubMed PMID:28356445.

[0485] 15. Anderson ST, Kaforou M, Brent AJ, Wright VJ, Banwell CM, Chagaluka G, et al. “Diagnosis of childhood tuberculosis and host RNA expression in Africa.” The New England journal of medicine. 2014;370(18):1712-23. doi:10.1056 / NEJMoa1303657. PubMed PMID: 24785206; PubMed Central PMCID: PMC4069985.

[0486] 16. Ramilo O, Allman W, Chung W, Mejias A, Ardura M, Glaser C, et al. “Gene expression patterns in blood leukocytes discriminate patients with acute infections.” Blood. 2007;109(5):2066-77. doi:10.1182 / blood-2006-02-002477. PubMed PMID:17105821; PubMed Central PMCID:PMCPMC1801073.

[0487] 17. Frangou EA, Bertsias GK, Boumpas DT. “Gene expression and regulation in systemic lupus erythematosus.” Eur J Clin Invest. 2013;43(10):1084-96. doi:10.1111 / eci.12130. PubMed PMID:23902282.

[0488] 18. Jia HL, Liu CW, Zhang L, Xu WJ, Gao XJ, Bai J, et al. “Sets of serum exosomal microRNAs as candidate diagnostic biomarkers for Kawasaki disease.” Scientific reports. 2017;7:44706. Epub 2017 / 03 / 21.doi:10.1038 / srep44706. PubMed PMID: 28317854; PubMed Central PMCID: PMCPMC5357789.

[0489] 19. Kuo HC, Hsieh KS, Ming-Huey Guo M, Weng KP, Ger LP, Chan WC, et al. “Next-generation sequencing identifies micro-RNA-based biomarker panel for Kawasaki disease.” The Journal of allergy and clinical immunology. 2016;138(4):1227-30. Epub 2016 / 07 / 28.doi:10.1016 / j.jaci.2016.04.050. PubMed PMID:27450727.

[0490] 20. Herberg JA, Kaforou M, Wright VJ, Shailes H, Eleftherohorinou H, Hoggart CJ, et al. “Diagnostic Test Accuracy of a 2-Transcript Host RNA Signature for Discriminating Bacterial vs Viral Infection in Febrile Children.” JAMA. 2016;316(8):835-45. Epub 2016 / 08 / 24. doi:10.1001 / jama.2016.11236. PubMed PMID:27552617.

[0491] 21. Hoang LT, Shimizu C, Ling L, Naim AN, Khor CC, Tremoulet AH, et al. “Global gene expression profiling identifies new therapeutic targets in acute Kawasaki disease.” Genome Med. 2014;6(11):541. doi:10.1186 / s13073-014-0102-6. PubMed PMID: 25614765. PubMed Central PMCID: PMCPMC4279699.

[0492] 22. Herberg JA, Kaforou M, Gormley S, Sumner ER, Patel S, Jones KD, et al. “Transcriptomic profiling in childhood H1N1 / 09 ​​influenza reveals reduced expression of protein synthesis genes.” J Infect Dis. 2013;208(10):1664-8. Epub 2013 / 08 / 01.doi:10.1093 / infdis / jit348. PubMed PMID:23901082; PubMed Central PMCID:PMCPmc3805235.

[0493] 23. Watanabe S. “Symptotical Equivalence of Bayes Cross Validation and Widely Applicable Information Criterion in Singular Learning Theory.” Journal of Machine Learning Research. 2010;11:3571-94.

[0494] 24. Johnson WE, Li C, Rabinovic A. “Adjusting batch effects in microarray expression data using empirical Bayes methods.” Biostatistics (Oxford, UK) 2007;8(1):118-27. Epub 2006 / 04 / 25.doi:10.1093 / biostatistics / kxj037. PubMed PMID:16632515.

[0495] 25. Abe J, Ebata R, Jibiki T, Yasukawa K, Saito H, Terai M. “Elevated granulocyte colony-stimulating factor levels predict treatment failure in patients with Kawasaki disease.” Journal of Allergy and Clinical Immunology. 2008;122(5):1008-13.e8. Epub 2008 / 10 / 22. doi:10.1016 / j.jaci.2008.09.011. PubMed PMID:18930517.

[0496] 26. Abe J, Jibiki T, Noma S, Nakajima T, Saito H, Terai M. "Gene expression profiling of the effect of high-dose intravenous Ig in patients with Kawasaki disease." Journal of immunology 2005;174(9):5837-45. PubMed PMID:15843588.

[0497] 27. Fury W, Tremoulet AH, Watson VE, Best BM, Shimizu C, Hamilton J, et al. “Transcript abundance patterns in Kawasaki disease patients with intravenous immunoglobulin resistance.” Hum Immunol. 2010;71(9):865-73. doi:10.1016 / j.humimm.2010.06.008. PubMed PMID: 20600450; PubMed Central PMCID: PMCPMC2929310.

[0498] 28. Popper SJ, Shimizu C, Shike H, Kanegaye JT, Newburger JW, Sundel RP, et al. “Gene-expression patterns reveal underlying biological processes in Kawasaki disease.” Genome Biol. 2007;8(12):R261. doi:10.1186 / gb-2007-8-12-r261. PubMed PMID:18067656; PubMed Central PMCID:PMCPMC2246263.

[0499] 29. Popper SJ, Watson VE, Shimizu C, Kanegaye JT, Burns JC, Relman DA. “Gene transcript abundance profiles distinguish Kawasaki disease from adenovirus infection.” J Infect Dis 2009;200(4):657-66. Epub 2009 / 07 / 09.doi:10.1086 / 603538. PubMed PMID:19583510; PubMed Central PMCID:PMC2878183.

[0500] 30. Ebihara T, Endo R, Kikuta H, Ishiguro N, Ma X, Shimazu M, et al. “Differential gene expression of S100 protein family in leukocytes from patients with Kawasaki disease.” European journal of pediatrics. 2005;164(7):427-31. doi:10.1007 / s00431-005-1664-5. PubMed PMID:15838637.

[0501] 31. Hu XR, Yu JS, Crosby SD, Storch GA. “Gene expression profiles in febrile children with defined viral and bacterial infection.” Proc Natl Acad Sci USA. 2013;110(31):12792-7. doi:10.1073 / pnas.1302968110. PubMed PMID:WOS:000322441500067.

[0502] 32. O'Hanlon TP, Rider LG, Gan L, Fannin R, Paules RS, Umbach DM, et al. “Gene expression profiles from discordant monozygotic twins suggest that molecular pathways are shared among multiple systemic autoimmune diseases.” Arthritis Res Ther. 2011;13(2):R69. doi:10.1186 / ar3330. PubMed PMID:21521520; PubMed Central PMCID:PMCPMC3132064.

[0503] 33. Ishii T, Onda H, Tanigawa A, Ohshima S, Fujiwara H, Mima T, et al. “Isolation and expression profiling of genes upregulated in the peripheral blood cells of systemiclupus erythematosus patients.” DNA Res. 2005;12(6):429-39. doi:10.1093 / dnares / dsi020. PubMed PMID:WOS:000242119900004.

[0504] 34. Fabriek BO, van Bruggen R, Deng DM, Ligtenberg AJ, Nazmi K, Schornagel K, et al. “The macrophage scavenger receptor CD163 functions as an innate immune sensor for bacteria.” Hematology 2009;113(4):887-92. doi:10.1182 / blood-2008-07-167064. PubMed PMID:18849484.

[0505] 35. Minich LL, Sleeper LA, Atz AM, McCrindle BW, Lu M, Colan SD, et al. “Delayed diagnosis of Kawasaki disease: what are the risk factors?” Pediatrics. 2007;120(6):e1434-40. doi:10.1542 / peds.2007-0815. PubMed PMID:18025079.

[0506] 36. Newburger JW, Takahashi M, Gerber MA, Gewitz MH, Tani LY, Burns JC, et al. “Diagnosis, treatment, and long-term management of Kawasaki disease: a statement for health professionals from the Committee on Rheumatic Fever, Endocarditis and Kawasaki Disease, Council on Cardiovascular Disease in the Young, American Heart Association.” Circulation 2004;110(17):2747-71. doi:10.1161 / 01.CIR.0000145143.19711.78. PubMed PMID:15505111.

[0507] 37. Petty RE, Southwood TR, Manners P, Baum J, Glass DN, Goldenberg J, et al. “International League of Associations for Rheumatology classification of juvenile idiopathic arthritis: second revision, Edmonton, 2001.” J Rheumatol. 2004;31(2):390-2. PubMed PMID:14760812.

[0508] 38. Du P, Kibbe WA, Lin SM. “lumi: an API for processing Illumina microarrays.” Bioinformatics. 2008;24(13):1547-8. doi:10.1093 / bioinformatics / btn224. PubMed PMID:18467348.

[0509] 39. Hoang LT, Shimizu C, Ling L, Naim AN, Khor CC, Tremoulet AH, et al. “Global gene expression profiling identifies novel therapeutic targets in acute Kawasaki disease.” Genome Medicine 2014;6(11):541. doi:10.1186 / s13073-014-0102-6. PubMed PMID: 25614765; PubMed Central PMCID: PMCPMC4279699.

[0510] 40. Herberg JA, Kaforou M, Gormley S, Sumner ER, Patel S, Jones KD, et al. “Transcriptome analysis of childhood H1N1 / 09 ​​influenza reveals reduced expression of protein synthesis genes.” J Infect Dis 2013;208(10):1664-8. Epub 2013 / 08 / 01.doi:10.1093 / infdis / jit348. PubMed PMID: 23901082; PubMed Central PMCID: PMCPmc3805235.

[0511] 41. Johnson WE, Li C, Rabinovic A. “Adjusting for batch effects in microarray expression data using empirical Bayesian methods.” Biostatistics (Oxford, UK) 2007;8(1):118-27. Epub 2006 / 04 / 25.doi:10.1093 / biostatistics / kxj037. PubMed PMID:16632515.

[0512] 42. Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. “limma powers differential expression analyses for RNA-sequencing and microarray studies.” Nucleic Acids Res. 2015;43(7):e47. doi:10.1093 / nar / gkv007. PubMed PMID:25605792; PubMed Central PMCID:PMCPMC4402510.

[0513] 43. Watanabe S. “Asymptotic equivalence of Bayesian cross-validation and widely used information criteria in singular learning theory.” Journal of Machine Learning Research, 2010;11:3571-94.

[0514] 44. Gelman A, Hwang J, Vehtari A. “Understanding predictive information criteria for Bayesian models.” Statistics and Computing. 2014; 24(6): 997-1016.

[0515] 45. Robin X, Turck N, Hainard A, Tiberti N, Lisacek F, Sanchez JC, et al. “pROC: an open-source package for R and S+ to analyze and compare ROC curves.” BMC bioinformatics. 2011;12:77. doi:10.1186 / 1471-2105-12-77. PubMed PMID: 21414208; PubMed Central PMCID: PMC3068975.

[0516] 46. ​​Youden WJ. "Index for rating diagnostic tests." Cancer. 1950; 3(1): 32-5. Epub 1950 / 01 / 01. PubMed PMID: 15405679.

[0517] 47. Bernardo JM, Smith AFM. Bayesian theory. Chichester, UK; New York: Wiley; 1994. Xiv, p. 586.

Claims

1. Use of one or more reagents for determining the expression level of a gene signature comprising PYROXD2 and at least four of the following genes from an RNA sample derived from a subject in the preparation of a product for identifying the subject as having Kawasaki disease (KD): CACNA1E, DDIAS, KLHL2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163, and RTN1.

2. The use according to claim 1, wherein the gene signature comprises PYROXD2 and 5, 6, 7, 8, 9, 10, 11 or 12 of the following genes: CACNA1E, DDIAS, KLHL2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163 and RTN1.

3. The use according to claim 1, wherein the gene signature comprises CACNA1E.

4. The use according to claim 1, wherein the gene signature comprises SMOX.

5. The method according to claim 1, wherein the gene signature comprises CD163.

6. The use according to claim 1, wherein the gene signature comprises: (i) PYROXD2 and CACNA1E; (ii) PYROXD2 and SMOX; or (iii) PYROXD2, CACNA1E and SMOX.

7. The use according to claim 1, wherein the gene signature comprises any one of the following gene combinations or consists of the gene combination: (i) PYROXD2, CACNA1E, CD163, KLHL2, and SMOX; (ii) PYROXD2, CACNA1E, IFI27, KLHL2, and SMOX; (iii) PYROXD2, CACNA1E, HS.553068, IFI27, and SMOX; (iv) PYROXD2, DDIAS, CACNA1E, IFI27, and SMOX; (v) PYROXD2, CACNA1E, CD163, KLHL2, and ZNF185; (vi) PYROXD2, DDIAS, CD163, KLHL2 and SMOX; (vii) PYROXD2, CACNA1E, CD163, IFI27, KLHL2, and SMOX; (viii) PYROXD2, CACNA1E, CD163, KLHL2, LINC02035, and SMOX; (ix) PYROXD2, DDIAS, CACNA1E, CD163, IFI27 and SMOX; (x) PYROXD2, CACNA1E, CD163, HS.553068, IFI27, and SMOX; (xi) PYROXD2, CACNA1E, CD163, KLHL2, SMOX, and ZNF185; (xii) PYROXD2, CACNA1E, IFI27, KLHL2, RTN1 and SMOX; (xiii) PYROXD2, CACNA1E, CD163, CLIC3, KLHL2 and SMOX; (xiv) PYROXD2, CACNA1E, CLIC3, IFI27, KLHL2, and SMOX; (xv)PYROXD2, DDIAS, CACNA1E, IFI27, RTN1 and SMOX; (xvi)PYROXD2, DDIAS, CD163, IFI27, KLHL2 and SMOX; (xvii) PYROXD2, CACNA1E, CD163, HS.553068, IFI27, KLHL2, and SMOX; (xviii) PYROXD2, CACNA1E, CD163, CLIC3, IFI27, KLHL2, and SMOX; (xix)PYROXD2, DDIAS, CACNA1E, CD163, IFI27, KLHL2 and SMOX; (xx)PYROXD2, CACNA1E, CD163, IFI27, KLHL2, RTN1 and SMOX; (xxi)PYROXD2, DDIAS, CACNA1E, CD163, HS.553068, IFI27 and SMOX; (xxii) PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2 and SMOX; (xxiii) PYROXD2, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, and SMOX; (xxiv) PYROXD2, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1, and SMOX; (xxv) PYROXD2, CACNA1E, CD163, HS.553068, IFI27, KLHL2, S100P, and SMOX; (xxvi) PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2 and SMOX; (xxvii) PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1 and SMOX; (xxviii) PYROXD2, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1, and SMOX; (xxix)PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1, S100P and SMOX; (xxx)PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1 and SMOX; (xxxi)PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1, S100P and SMOX; (xxxii) PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, LINC02035, RTN1, S100P and SMOX; (xxxiii) PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, RTN1, S100P, SMOX and ZNF185; (xxxiv)PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, RTN1, S100P, SMOX and ZNF185; (xxxv) PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, HS.553068, IFI27, KLHL2, LINC02035, RTN1, S100P, and SMOX; or (xxxvi)PYROXD2, DDIAS, CACNA1E, CD163, CLIC3, IFI27, KLHL2, LINC02035, RTN1, S100P, SMOX, and ZNF185.

8. The use according to claim 1, wherein the gene signature comprises or consists of: CACNA1E, DDIAS, KLHL2, PYROXD2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163 and RTN1.

9. The use according to any one of claims 1 to 8, wherein the use further comprises detecting the expression level of one or more housekeeping genes selected from the group consisting of actin, GAPDH, ubiquitin, 18s rRNA, RPII (POLR2A), TBP, PPIA, GUSB, HSPCB, YWHAZ, SDHA, RPS13, HPRT1 and B4GALT6.

10. The use according to any one of claims 1 to 8, wherein a subject suffering from KD can be identified in the presence of one or more of the following or distinguished from patients suffering from one or more of the following: a bacterial infection, a viral infection and an inflammatory condition.

11. The use according to any one of claims 1 to 8, wherein the subject is a child in the age range of 2 to 59 months.

12. The use according to any one of claims 1 to 8, wherein the subject has a fever.

13. The use according to any one of claims 1 to 8, wherein the analysis of gene expression regulation employs microarray, gene chip, PCR, RT-PCR or multiplex PCR.

14. Use of a primer set or a gene chip comprising probes specific to PYROXD2 and at least four of the following genes in the preparation of a product for detecting Kawasaki disease (KD) in a sample: CACNA1E, DDIAS, KLHL2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163 and RTN1, wherein the primer set comprises primers specific to polynucleotide gene transcripts from PYROXD2 and at least four of the following genes: CACNA1E, DDIAS, KLHL2, SMOX, ZNF185, LINC02035, CLIC3, S100P, IFI27, HS.553068, CD163 and RTN1.

Images