Diagnostic test for the detection of bladder cancer in urine

Abstract

It is provided a method of diagnosing bladder cancer in a subject comprising measuring in a urine sample the protein level of expressions of at least two protein markers selected from complement factor H (CFH), Fibrinogen β (FGB), alpha-2-macroglobulin (A2M) and pancreatic alpha-amylase (AMY2A) and a kit for same. The method and kit allow to monitor and / or diagnose high-grade bladder cancer in patients suspected of having bladder cancer, having or not hematuria.

Description

DIAGNOSTIC TEST FOR THE DETECTION OF BLADDER CANCER IN URINECROSS-REFERENCE TO RELATED APPLICATION

[0001] The present application is claiming priority from U.S. Provisional Application No. 63 / 661 ,715 filed June 19, 2024, the content of which is hereby incorporated by reference in its entirety.TECHNICAL FIELD

[0002] It is provided a method of diagnosing bladder cancer in a patient using a urine sample.BACKGROUND

[0003] Bladder cancer is a worrying public health issue because of its high incidence since it is the 5thmost common cancer and generates 12,000 new cases per year, but also because of the financial burden it imposes on the health system. Indeed, it is the most expensive cancer to treat, among other things because 80% of patients who obtain a first diagnosis recur within 2 years. This makes the level of surveillance of this disease very high and many patients will undergo several cystoscopies for many years.

[0004] Cystoscopy accompanied by urinary cytology are diagnostic tools that can detect the presence of tumors in the bladder. These are tools used for several decades which have ensured the role of screening for bladder tumors on a global scale. However, they remain imperfect and have numerous limitations, such as the comfort level of the procedure, detection performance and associated costs. The main symptom associated with this pathology is the presence of blood in the urine, a health condition whose incidence can vary between 2 to 5% in the general population. Patients with some form of hematuria are commonly referred to the urology department for further evaluation of their condition. As a result, waiting lists are growing and no patient benefits from preliminary triage that can prioritize according to the severity of the tumor form, if applicable.

[0005] It is thus highly desired to be provided with means that could predict the presence of bladder cancers in a non-invasive, less painful and less expensive way for monitoring patients with suspected bladder lesions.SUMMARY

[0006] It is provided a method of diagnosing bladder cancer in a subject comprising providing a urine sample from the subject; and measuring level of expressions of at least two protein markers selected from the group consisting of complement factor H (CFH), Fibrinogen p (FGB), alpha-2-macroglobulin (A2M) and pancreatic alpha-amylase (AMY2A).

[0007] In an embodiment, the method comprises measuring the level of expression of CFH and FGB.

[0008] In another embodiment, the method comprises measuring the level of expression of CFH and A2M.

[0009] In another embodiment, the method comprises measuring the level of expression of CFH and AMY2A.

[0010] In another embodiment, the method comprises measuring the level of expression of FGB and A2M.

[0011] In another embodiment, the method comprises measuring the level of expression of FGB and AMY2A.

[0012] In another embodiment, the method comprises measuring the level of expression of A2M and AMY2A.

[0013] In another embodiment, the method comprises measuring the level of expression of CFH, FGB and A2M.

[0014] In another embodiment, the method comprises measuring the level of expression of CFH, A2M and AMY2A.

[0015] In another embodiment, the method comprises measuring the level of expression of FGB, A2M and AMY2A.

[0016] In another embodiment, the method comprises measuring the level of expression of CFH, FGB, A2M and AMY2A.

[0017] In a further embodiment, an increase in expression levels of CFH, FGB orA2M is measured, or a decrease in expression level of AMY2A is measured.

[0018] In an embodiment, the expression level of the at least two protein markers is measured using an antibodies, nanobodies or aptamers.

[0019] In an embodiment, the expression level of at least two protein markers is measured by contacting an antibody binding to CFH, FGB, A2M or AMY2A to the urine sample.

[0020] In a further embodiment, the expression level of at least two protein markers are normalized to a reference value.

[0021] In a further embodiment, the level of expressions of at least two protein markers are measured using an ELISA test, a mass spectrometry analysis or a rapid test.

[0022] In an embodiment, the urine sample is from a patient with hematuria.

[0023] In a further embodiment, the urine sample is from a patient in remission.

[0024] In another embodiment, the bladder cancer is a high-grade cancer.

[0025] It is further provided a kit for diagnosing bladder cancer in a subject comprising at least two detecting molecules binding to at least two proteins selected from complement factor H (CFH), Fibrinogen p (FGB), alpha-2-macroglobulin (A2M) and pancreatic alpha-amylase (AMY2A), and instruction for use.

[0026] In an embodiment, the detecting molecules are antibodies, nanobodies or aptamers.

[0027] In a further embodiment, the kit further comprises a filter paper to collect a urine sample from said subject.

[0028] In an embodiment, the kit is an ELISA kit, a mass spectrometry kit or a rapid test.BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Reference will now be made to the accompanying drawings.

[0030] Fig. 1 illustrates LFQ intensity data normalization with quantile classification method. X-axis represents the order in which samples were analyzed on the mass spectrometer from the 1st to the 388th (97 urine samples x 1 technical duplicate(aliquots) = 194 samples x 1 mass spectrometer duplicate = 388 samples). The different points stand for the different periods where the samples were analyzed on the mass spectrometer and therefore allows to distinguish the different batches. The Y-axis corresponds to the mean LFQ intensity (Log2) calculated for each sample. The two graphs show the samples distribution before in (A) and after (B) normalization with the quantile classification method.

[0031] Fig. 2 illustrates the Principal Component Analysis (PCA) on urine samples proteomic data. Each dot represents a urine sample (4 samples per patient) and its x and y coordinates indicate the first two principal components (PC1 , PC2) and its associated variance. The effect of in (A) the MS batch (1 , 2 or 3) or in (B) the technical replicate for each patient’s urine sample (Aliquot A or B). The dots vary in function of the MS batch effect (A) or the aliquot effect (B).

[0032] Fig. 3 illustrates PCA of four biological variables characterizing bladder cancer on the urinary proteome of samples from the discovery cohort. The principal component analysis presented above reveal the two principal directions (X- and Y-axis) where the biggest gap between data can be explained in function of in (A) the tumor status; (B) the bladder cancer history; (C) the level of progression; and in (D) the level of risk associated to the degree of tumoral infiltration. Each dot represents a patient and its x and y coordinates indicates the variance explained by PC1 (X-axis) and PC2 (Y-axis). The NA data in graph of B, C and D may influence the number of visible dots of one graph to another.

[0033] Fig. 4 illustrates PCA of other biological variables known to modulate urinary proteome composition and abundance. The principal component analysis presented above reveals the two principal directions (X- and Y-axis) where the biggest gap between data can be explained in function of (A) the presence or not of chronic renal insufficiency; (B) the sex and (C) the smoking status. Each dot represents a patient and its x and y coordinates indicates the variance explained by PC1 (X-axis) and PC2 (Y-axis). The NA data in graph A and C may influence the number of visible dots of one graph to another.

[0034] Fig. 5 illustrates Volcano-plot displaying the difference in protein abundance between urine samples from patients with positive and negative bladder cancer status at the time of sampling. X-axis stands for Log2-ratio of normalized LFQ data and Y-axis stands for the level of statistical significance (p-values) determined by the non-parametric Mann-Whitney U test. Dots representing significantly depleted proteins in urine ofpatients with positive bladder cancer status, and dots representing significantly enriched proteins in urine of the same group of patients, are shown - compared to urine of patients with a negative bladder cancer status - independently of cancer stage, grade and previous cancer history. Black dots are proteins whose abundance is not significantly modulated in function of bladder tumor status. A value of 2 on the Y-axis corresponds to a p-value of 0.01.

[0035] Fig. 6 illustrates Volcano-plot displaying the difference in protein abundance between urine samples from patients with a negative status of bladder cancer at the time of sampling - comparing patients in remission of bladder cancer and patients with no history of bladder cancer status. X-axis is normalized LFQ Log2-ratio data and Y-axis is the level of statistical significance (p-value) determined by the non-parametric Mann- Whitney U test. The first group is made of patients with no history of bladder cancer (N) and the second is made of patients with a history of bladder cancer (RE).

[0036] Fig. 7 illustrates differential analysis of the urinary proteome for HGBC and LGBC using Volcano-plots. X-axis stands for Log2-ratio of normalized LFQ data and Y- axis stands for the level of statistical significance (p-value) determined by the nonparametric Mann-Whitney U test. Dots representing depleted proteins or enriched proteins in urine of patients are shown with (A) low-grade bladder cancer (LGBC) and (B) high-grade bladder cancer (HGBC) compared to urine of patients with a negative bladder cancer status at the moment of sampling - both having a history of bladder cancer or not (NRE). Black dots stand for proteins which are not significantly modulated in presence of bladder cancer, whether low-grade or high-grade. A value of 2 on the Y- axis corresponds to a p-value of 0.01. Proteins with a significant change in abundance of p<0.0001 are identified in white boxes closed to their corresponding dot.

[0037] Fig. 8 illustrates examples of results obtained following the Western Blot quantification of A2M, FGB, CFH and AMY2A. X-axis is the normalized intensity relative to a reference sample. The reference sample has both a negative and no history of bladder cancer status and comes from the discovery cohort. Its intensity has been adjusted to 100. The sample comes from the discovery cohort and has a HGBC status. The optimal threshold (in %) is shown to detect HGBC based on the ROC curves data. The number above the histogram bands are the exact normalized relative intensity values.

[0038] Fig. 9 illustrates ROC curves of Western blot quantification of the four biomarkers combination for the detection of HGBC. X-axis stands for false positives (100- specificity) whereas Y-axis stands for the sensitivity - the proportion of true positives. The red dotted line on each graph is the reference point where the AUC is 0.5 (random). Each AUC value for each protein is indicated in the lower right corner of each graph.DETAILED DESCRIPTION

[0039] It is provided a method of diagnosing bladder cancer in a subject comprising providing a urine sample from the subject, measuring level of expressions of at least two protein marker selected from the group consisting of complement factor H (CFH), Fibrinogen p (FGB), alpha-2-macroglobulin (A2M) and pancreatic alpha-amylase (AMY2A).

[0040] Accordingly, it is disclosed a urine test that predicts the presence of bladder cancer between an individual without cancer and an individual with high-grade cancer (aggressive cancer). The false positive rate varies depending on the detection threshold that is chosen, i.e. the minimum number of biomarkers whose intensity is greater than the detection threshold set by a ROC curve. The ability to predict the presence of cancer between an individual without cancer and an individual with low-grade cancer (non- aggressive cancer) is 55% considering a threshold of >2+ / 4. A cancer-free individual may be an individual who has never had bladder cancer or an individual in remission, i.e. who has had low- or high-grade bladder cancer in the past but who is currently without recurrence during sampling.

[0041] The use of the combination of the four protein biomarkers (FGB, A2M, CFH and AMY2A) present in urine as described herein allows to detect the presence of tumors in patients at high risk of recurrence and progression.

[0042] Factor H is a glycoprotein with an approximate molecular weight of 155 kDa that plays a role as a complement inhibitor. This protein produced by the liver as well as several other cell types such as fibroblasts, endothelial cells, monocytes, is found at a blood concentration varying between 200 to 800 pg / ml (Parente et al., 2017, Cellular and Molecular Life Sciences, 74(9), 1605-1624). Its binding to the major histocompatibility complexes, i.e. MHC-I, prevents complement activation and the amplification of these complexes on the cell surface. The lack of effective recognition by Factor H of host cells due to mutations and / or polymorphisms has been associated with several diseases and tissue damage linked to inadequate complement activation (Ferreira, Pangburn, et al.,2010, Molecular Immunology, 47(13)). CFH protein is also the main molecule used in an ELISA-type urine test marketed for the diagnosis of bladder cancer (Kinders et al., 1998, Clinical Cancer Research, 4(10), 2511-2520). However, this test is not used clinically, in particular because of its low sensitivity and specificity.

[0043] The FGB gene encodes the beta chain of the fibrinogen protein, a 420 kDa protein composed of three main chains: alpha (a), beta (P) and gamma (y). This protein is involved in the mechanisms of coagulation and of inflammation. Fibrinogen p chain has been previously identified as a potential biomarker for detecting bladder cancer (Linden et al., 2012, Proteomics, 12: 135-144).

[0044] The A2M protein is a 720 kDa plasma protein formed by four 180 kDa subunits whose main function is to bind peptides or self-proteins or which are of pathogenic origin. Its main role is that of a broad-spectrum protease inhibitor. To this end, A2M participates in the humoral and cellular immune defense system. One of the most notable publications about the A2M protein mentions its overexpression in cancer cells, particularly in melanoma (Borth, 1992, The FASEB Journal, 6(15), 3345-3353).

[0045] Pancreatic alpha-amylase is a pancreatic enzyme whose main function is to digest polysaccharides. Little is known about gene and protein dysregulation in cancer. However, a study that carried out large-scale genomic profiling demonstrated that several samples from patients suffering from gastric tumors present homozygous deletions in the chromosomal region where the AMY2A locus is located (Kang et al., 2010, International Journal of Oncology, 36, 1429-1435.).

[0046] Table 1 describes the two cohorts of participants based on a range of clinical variables. Since the individuals were mainly selected during cystoscopy consultation days, the study population reflects well the distribution of the profiles of typical patients seen in consultation and are representative of the clinical reality in terms of cancer screening bladder. Variables, like age and sex are distributed in a manner equivalent to what is listed in the literature (Berdik, 2017, Nature, 551 (7679), 34-35).Table 1 : Presentation of clinical and pathological risks data associated with participants included in both cohorts

[0047] In order to identify molecules which have real clinical potential for non- invasive diagnosis of bladder cancer, the data obtained following the analysis by mass spectrometry had to be filtered and processed rigorously using an analysis protocol and statistical interpretation in order to ensure efficiency of the targets identified.

[0048] The quantile normalization method made it possible to minimize the bias induced by the analysis method and the device. Each urine sample (2 samples / participant) was analyzed twice on the mass spectrometer in order to evaluate the technical reproducibility of the device (n=2, duplicate run). However, there are three distinct slots during which the samples were analyzed (see Fig. 1). The first occurrenceat which part of the samples was analyzed is described with the acronym Lot_MS 1 . Lot_MS 2 and Lot_MS 3 correspond to the other two occurrences. It is possible to observe a slight increase in the average LFQ intensity for samples that passed through the Lot_MS 3 compared to samples from the first two lots. A possible explanation that can justify this slight difference is the impact of the renewal of the column of separation which, freshly installed, is cleaner and contains fewer contaminants. Normalization by the quantile method referred to by Cuklina in the development of the R package proBatch 1.8.0 makes it possible to correct the bias induced by this technical aspect (Fig. 1 B) (Cuklina, 2018. Computational challenges in biomarker discovery from high- throughput proteomic data. ETH Zurich - Doctoral Thesis').

[0049] Once the intensity data have been normalized, a principal component analysis (PCA) was carried out to determine if the type of batch (1 , 2 or 3) could explain the variance between samples. Principal component analysis is a statistical analysis which makes it possible to reduce a complex and multidimensional data set under a limited number of principal components. This allows to quickly see if a variable such as gender or age, for example, has a significant impact on the composition of the samples The first two principal components are used to visualize the distribution of points (i.e. samples) in the form of a two-dimensional graph. Depending on the position of the points in space, it is possible to note the degree of similarity or, conversely, difference that exists between the samples (Ivosev et al., 2008, American Chemical Society, 80( 3), 4933- 4944). The x-axis and y-axis of Fig. 2A indicate that the first two principal components (PC1 and PC2) explain respectively 23% and 10% of the sample variance. This latter variance is not caused by the type of batch in which the samples were located at the time of their passage on the mass spectrometer. Their analyses were carried out on all samples analyzed by LC-MS / MS, for a total of 388. Distribution samples (and the batch with which they are associated) are considered homogeneous. Thus, the effect of the batch on the data is negligible following quantile normalization. Fig. 2B, for its part, illustrates the degree to which sample preparation impacts urinary proteome variance. The technical replicates, i.e. aliquots “A” and “B”, are distributed homogeneously suggesting that sample preparation is not a significant source of variability in the urinary proteome.

[0050] Hierarchical grouping associated with the heatmap is a statistical visualization tool capable of representing complex and multidimensional dataset by integrating it into a simple and understandable visual format. In order to evaluate generally the similarityof samples across their entire proteome, a Hierarchical clustering, using the k-means method, was carried out.

[0051] When collecting clinical data from patients who provided a urine sample, descriptive data relating to grade, stage, tumor history as well as the level of risk concomitant with the degree of tumor invasion were included in the analysis to evaluate their impact on the modification of the urinary proteome. It is important to specify that the principal component analysis is done on the same dataset. Thus, the distribution of points in two-dimensional space does not change. What changes are the parameters put in relation. In Fig. 3, the four different PCAs illustrate the distribution of samples according to different variables, namely, tumor status, history of bladder cancer, stage of cancer as well as the level of risk depending on the degree of tumor invasion. In Fig. 3A, the difference between patients suffering from bladder tumors, without considering stage, and patients without bladder tumor at the time of sampling, without considering tumor history, are highlighted. According to this graphic representation, it is not possible to clearly distinguish two groups of points in function of tumor status. This suggests that the composition and abundance of the urinary proteome is not globally influenced by tumor status. The PCA test evaluates the overall variance, meaning that the possibility of a subgroup of proteins being modulated by tumor status is entirely plausible. Figs. 3B, C and D only target patients with bladder cancer at time of sampling (n=44), considerably reducing the dataset. Figs. 3B and 3C do not seem to demonstrate the formation of groups of points defined according to the parameters observed. The participants (points in the graph), and therefore the categories to which they are associated, are distributed fairly homogeneously in the two-dimensional space.

[0052] This suggests that the history (primary tumor or recurrence) and stage of progression are not variables that explain the distance observed between points. As for Fig. 3D, it can be observed that the participants belonging to the category of non-invasive cancers (low risk) - exclusively low grade - are all found in the upper part of the graph (PC1 > -0.1). It is not possible to draw a firm conclusion, however the results suggest that the level of tumor risk partly explains the variability of the urinary proteome of patients with tumors. In addition to the biological variables directly linked to the pathology, other biological variables deserve to be considered in the analysis in order to evaluate their contribution to the variability of the urinary proteome of the participants part of the study. Fig. 4 shows the three additional biological variables for which a Principal Component Analysis has been carried out.

[0053] Data from the 97 participants in this cohort were compared according to glomerular filtration rate, tobacco consumption and sex of the individual. These results suggest that no variable has a notable effect on the variance of the urinary proteome of the participants. This approach allows to determine what type of variable, whether biological, technical or residual, influences the data variance. It was determined that 70% of the data variance of the urinary proteome was observed and is explained by biological variables. 5% of the data variance is attributed to the batch effect when it is linked to GFR, tumor status as well as the stage of progression of the cancer. In other words, a weighted relative average variance of 5% is explained by the batch effect when considering data according to tumor status, stage of progression and GFR. Finally, 25% of the variance is explained by residual variables, meaning they are not clearly defined.

[0054] In order to determine which elements of the urinary proteome were significantly dysregulated in the presence of bladder cancer, a non-parametric test of Mann-Whitney U for unpaired data was carried out. Moreover, the test was accompanied by the calculation of differential expression ratios for each protein detected by comparing the urine of bladder cancer negative versus positive participants at the time of sampling. Fig. 5 is the graphic representation of the volcano-plot of this mathematical operation. This figure allows us to observe the proportion of proteins which undergo significant alterations (p<0.01) in terms of abundance when comparing the two conditions. Fig. 5 illustrates higher levels of depleted proteins than enriched proteins in patients with bladder cancer. The presence of tumor(s) in the bladder seems to increase the concentration of a subgroup of proteins significantly. The grade, a histopathological criterion reflecting the degree of aggressiveness of the tumor, is a biological variable which was integrated into the analyses in order to determine whether the latter could have a significant impact on the composition and variation in abundance of the urinary proteome. A differential analysis of protein abundance was carried out on all participants without bladder cancer at the time of sampling to assess whether tumor history had a significant impact on the distribution of the urinary proteome in terms of the intensity of its components.

[0055] Fig. 6 presents the results obtained in the form of a volcano-plot and confirms that the urinary proteome varies very little between patients in remission of a bladder tumor, whether the tumor had a low or high grade, when compared to patients with no history of bladder tumor. Considering this homogeneity of samples, a single group, named NRE, was formed with all participants without bladder tumors in order to facilitateanalyses. In Fig. 7, the two graphs make it possible to differentiate the urinary proteome of patients suffering from high-grade ( HGBC group) and low-grade (LGBC group) tumors, with the urinary proteome of patients in the tumor-free group (NRE). By looking at the graphs panels A and B, a marked difference between the proteome of high-grade patients and that of low-grade patients is observed. When comparing urine of group HGBC and those of group LGBC, it is possible to observe a greater enrichment of certain proteins. These proteome alterations suggest specificity of the latter at the level of the tumor grade at the time of sampling.

[0056] Once the MS data was cleaned and normalized, a first list of proteins was identified following performance in volcano plots and the non-parametric Mann-Whitney test on the different groups of patients. The selection criteria for the quality of the MS identifications were applied, which led to the selection of a shortened list of proteins. The ROC curves were carried out. A final list of 4 proteins was determined representing those for which best results by Western Blots (WB) was obtained and which provided better sensitivity and specificity when combined.

[0057] The current sensitivity and specificity of cystoscopy for the detection of bladder cancer is 75% and 72%, respectively. To be able to convince urologist associations to accept the changes in their practice, it is estimated that one need to provide a clinical prediction algorithm that has a sensitivity of at least 85% and a specificity of at least 82% (10% higher). Based on this goal, the sample size of 1001 (501 cancer, 500 control) for the current development dataset meets the recommendations for the development of a clinical prediction model for binary outcomes, which are to ensure the sample size is enough to (1) estimate the overall outcome proportion with sufficient precision(>=385); (2) target a small mean absolute prediction error (MAPE) (>= 830, assuming a MAPE of 0.05 and no more than 12 predictors); (3) target a shrinkage factor of 0.9 (>= 478, assuming that the Cox-Snell RA2 reaches at least 0.2 and that no more than 12 predictors are used) and (4) target small optimism of 0.05.

[0058] Considering that the algorithm is expected to have a sensitivity of at least 93.05% based on the current values, a sample of 99 cancer patients will enable the detection of a statistically significant 10% difference with the sensitivity of cystoscopy with a power of 80% under a two-sided test for a single proportion based on the normal approximation to the binomial distribution, at level alpha=5%. Furthermore, considering that the algorithm is expected to have a specificity of at least 90.85%, a sample of 100 control patients will enable the detection of a statistically significant 10% difference withthe specificity of cystoscopy with a power of 80% under a two-sided test for a single proportion based on the normal approximation to the binomial distribution, at level alpha=5%.

[0059] Once the final list of antibodies is selected, an evaluation of performance based on each protein analysis was carried out to discriminate urine samples coming from healthy patients and those coming from bladder cancer patients. The Receiver Operating Characteristics (ROC) curves are used to determine the numerical threshold for which the sensitivity and specificity of detection is optimal. This threshold value can therefore correspond to the protein concentration when used as a biomarker.

[0060] After some refinements, the antibodies targeting the A2M, FGB, CFH and AMY2A have demonstrated their effectiveness in binding proteins of interest (Fig. 8) serves as an example). It is provided that the signal detected in urine samples from highgrade tumor patients is significantly higher than that of the other samples, with regard to the A2M, FGB and CFH proteins. On the other hand, the signal detected for the AMY2A protein is higher in samples from patients with absence of bladder tumor and that, regardless of tumor history. The urine samples from patients with bladder tumors at the time of sampling, whether low-grade or high-grade, presents a weaker signal for this protein. There is a significant difference between the different groups of urine samples for all four proteins measured thanks to the use of the non-parametric multiple comparison test Kruskal-Wallis. A p-value less than 0.0001 characterizes the difference between the high-grade patients and the ones not suffering from high-grade tumors, i.e. NRE and LGBC patients for A2M, FGB and CFH proteins. Indeed, no significant difference is measured when comparing samples from the NRE and LGBC patients. Concerning the AMY2A protein, a significant difference between NRE patients and CA3 patients was measured.

[0061] Fig. 9 presents the ROC curves obtained using the data from WB experiments. An average AUC value of 0.87 was reached for all 4 proteins, compared to 0.82 for the same proteins identified by LC-MS / MS. In general, the values remain relatively similar except for AMY2A which presents a notable improvement, going from an AUC value of 0.75 to a value of 0.86. Threshold values identified in the squares represent the normalized relative intensity where the number of true positives (sensitivity) is maximized relative to the number of false positives (specificity). Thus, for the A2M protein, a minimum relative normalized intensity 5 times higher than the reference value, i.e. a sample without a typical bladder tumor whose value has been set to 1 , allowsoptimal detection of samples coming from patients with high-grade bladder tumor with a sensitivity of 87% and a specificity of 80%. As for the FGB and CFH protein, a relative normalized intensity 1 .6 times and 1 .5 times higher than the reference makes it possible to detect patients with high-grade tumors with a sensitivity of 96% and 87% and a specificity of 72% and 77%, respectively. As for the AMY2A protein which is detected at a lower level in urine samples from patients with bladder tumors, a relative normalized intensity equivalent to 0.65 times the reference value or less provides a sensitivity of 83% and a specificity of 83% as well.

[0062] Fig. 8 illustrates different representative examples of the results obtained following the validation stage of potential biomarkers, identified during the investigation stage by mass spectrometry, on the 104 urine samples from the independent cohort. Around 10 WBs were carried out by antibody: the results obtained on a membrane are presented for each antibody. The different tumor status associated with samples are not indicated in the figure because the objective is to evaluate the signal obtained following the revelation of the secondary antibody which recognizes the A2M, FGB, CFH and AMY2A. The line marks the threshold allowing the samples to be identified “positive” for a high-grade bladder tumor. These threshold values were determined using the results obtained by the ROC curves (Fig. 9).

[0063] Different values associated with the performance of biomarkers used individually or in combination, distinguishing patients without tumors according to their tumor history, i.e. without history of bladder tumor or in remission of a bladder tumor, were determined (see Table 2).Table 2: Performances of biomarkers identified in distinguishing patients with or without tumors

[0064] The use of the combination makes it possible to improve the intake of each individual biomarker in order to create a highly sensitive and specific tool for the detection of tumors at high risk of recurrence or progression. Indeed, when a urine sample is analyzed and at least two biomarkers on four obtain a normalized relative intensity value above the threshold, if they are overexpressed biomarkers (CFH, FGB or A2M), or below the threshold, if it is AMY2A biomarker, then the sample is considered positive for a high grade bladder cancer. Indeed, the identity of the two positive biomarkers does not influence test success: each biomarker has equal weight in the combination. So, there are 11 possible combinations allowing the identification of high-grade bladder tumors in urine. The urine of patients who have already suffered from a bladder tumor, whether it was low- or high-grade, shows a slightly higher number of false positives rather than for the patients with no history, i.e. 27% compared to 20% respectively.

[0065] In order to evaluate whether the combination of the four biomarkers could compete in terms of performance with the non-invasive method currently recommended by the Canadian Urological Association for the detection of high-grade tumors, i.e. urine cytology, a comparative study of urine cytology with the combination of the four biomarkers quantified by WB was carried out. Firstly, the evaluation of the performance of cytology was evaluated on the data of 233 patients integrated into one or the other study cohorts. As for the comparative analysis, only the data of the participants included in the second cohort were evaluated, i.e. the participants whose urine samples wereanalyzed by WB. Results aimed at demonstrating the performance of urine cytology based on local data, from patients residing in the region and having been analyzed by CHUS pathologists, was first obtained. The objective of this first analysis was to see if the local data agreed with those presented in the literature for the performance of this method for bladder cancer detection. Data from 233 urine cytologies carried out between 2015 and 2020 were collected on ARIANE, the CHUS local patients database. A distinction is made between “positive and atypical” cytologies and those which are “positive” only. Indeed, histopathology reports describe cytologies either as being benign, atypical, suspicious or positive. Benign cytology means that no dysplastic cells have been found. Atypical cytology can be the result of a myriad of possible causes, such as an infection or chronic illness (other than bladder or kidney), to name a few. Thus, urologists do not immediately consider atypical urinary cytology as an alarming sign of bladder cancer. However, suspicious or positive cytologies are considered true positives and patients are seen for cystoscopy quickly.

[0066] Therefore, to evaluate the performance of actual urine cytology, only positive cytologies should be taken into account. However, including atypical cytologies in the analyses makes it possible to evaluate quantitatively whether or not urologists would benefit from considering them in their diagnostic decision-making algorithm. The low- grade bladder tumors, considered at low risk of progression and recurrences, are very rarely detected by urine cytology. The sensitivity obtained when including the results of atypical cytologies is 37% and 9% when only positive cytologies are included (see Table 3). An average increase of 25% in sensitivity is observed for the detection of high-grade tumors. However, it remains less than 40% when considering positive cytologies only. The specificity related to the detection of false positives is 87% in patients with no history of bladder cancer and remains the same, whether atypical cytologies are included or not. For patients in remission of a low-grade or high-grade tumor, a difference exists at the test level regarding the detection of false positives (specificity) when atypical cytologies are included. Generally speaking, urine cytology remains true to itself by presenting excellent specificity, particularly at the level of high grade tumors, and a very low sensitivity for the detection of low-grade tumors and, surprisingly, for the detection of high-grade bladder tumors.Table 3: Analysis with samples from discovery and validation cohorts

[0067] A comparison of performance between urine cytology and the combination of the four biomarkers quantified by WB for all of the 104 samples making up the validation cohort was made. The groups of patients suffering the most from hematuria are those who have no history of bladder tumor and patients with high grade tumors. The average rate of hematuria for these groups, macroscopic and microscopic combined, exceeds 70%. Generally speaking, the main difference between urine cytology and the combination of biomarkers is the almost opposite kind of performance in terms of their respective sensitivity and specificity. Indeed, cytology is poorly sensitive unlike the combination of biomarkers which achieves perfect sensitivity for all high-grade tumors, all stages combined, on the study patients. It is only when applying a threshold of minimum two positive biomarkers out of four that specificity becomes relevant for diagnostic purposes. As for the specificity of urine cytology, it has perfect values, i.e. 100%, for all categories of patients. This involves evaluating concretely the effectiveness of these two methods in targeting the most at risk patients, among a pool of patients without a history of tumor at the time of sampling. If only high-grade participants are considered, the sensitivity of positive urine cytology varies between 33 and 50%. Thus, more than 50% of high-risk participants are missed. On the other hand, the algorithm using a combination of four biomarkers demonstrated in this study identifies 100% of patients suffering from a high-grade bladdertumor. From participants suffering from highgrade non-invasive tumors (n=9), 60% also suffered gross hematuria at the time of sampling. 25% of participants with high-grade invasive bladder tumors (n=4) also had macroscopic hematuria.

[0068] It is thus encompassed measuring level of expressions of at least two protein markers selected from the group consisting of complement factor H (CFH), Fibrinogen p (FGB), alpha-2-macroglobulin (A2M) and pancreatic alpha-amylase (AMY2A) using e.g. but not limited to, an antibodies, nanobodies and aptamers.

[0069] In an embodiment, the level of expressions of the at least two protein markers are measured using an ELISA test, a mass spectrometry analysis, or a rapid test such as a test strip. Rapid testing is a known method for qualitatively and quantitatively analyzing a trace amount of analyte in a short period of time. An assay strip or test strip comprises a reaction substance capable of reacting with a target substance to be detected by using an antigen-antibody reaction or an analyte strip containing the reaction substance. Typically the presence of the analyte is visualize by a change of color on the strip.

[0070] As provided herein, two detection methods are encompassed as a diagnostic test using the combination of the four proteins. Initially, the development of an ELISA type test is possible knowing that the four proteins are relatively abundant, were detected by WB and present a marked difference in abundance depending on the tumor status. The quantitative mass spectrometry approach called Parallel Reaction Monitoring (PRM) is an alternative that can easily be implemented in hospital clinical biochemistry laboratories. This technique requires lower initial costs when the mass spectrometer is already available and the acquisition of AQUA peptides, i.e. a synthetic tryptic peptide serving as a reference marker for the proteins to be detected, is simple and rapid. Furthermore, in terms of difficulty of development, the PRM and the ELISA are quite equivalent. The PRM is a method allowing to increase the number of targets to detect, at little additional cost, compared to the ELISA approach. However, although this may provide some advantages, it is not necessarily desirable to have a combination with a high number of biomarkers, as this can significantly reduce the specificity of the test. Consideration is required regarding the logistics of transport and storage of the diagnostic test.

[0071] The present demonstration was done on fresh urine samples, placed immediately on ice and stored for long-term storage in -80°C freezers. In the event that such a test is integrated into clinical practices and the analyses are centralized in a single hospital center for example. The diagnostic test described herein can be adapted to be used as a simple filter paper(FP) that can be manipulated when urinating. The filter paper (FP) would be transported or sent by post, without additional constraints in terms of bothconservation and transport. These FPs can be analyzed in a clinical biochemistry laboratory, whose facilities can be centralized. An ELISA-type screening device, similar to regular over-the-counter pregnancy tests, is also encompassed.

[0072] The majority of new consultations referred to the urology department are for problems of macroscopic or microscopic hematuria whose cause is unknown. The group of patients without tumor (N) well represents the clinical reality in terms of proportion of cases of hematuria seen in urology, i.e. more than 70% of new consultations are linked to a problem of hematuria. It is important to remember that the proportion of people suffering from hematuria concomitantly with a bladder tumor is 10% for macroscopic hematuria and 2 to 5% for microscopic hematuria. Thus, since a significant number of people suffer from this condition, but few of them also suffer from a bladder tumor, the need to stratify these patients according to the risk they present is important. The method provided herewith can thus be used to screen and stratify patients having problems of macroscopic or microscopic hematuria.

[0073] Another application for the provided test as encompassed herein is its postcystoscopy use for patients suffering from macroscopic hematuria; case where the patient is directly prioritized for cystoscopy. Generally speaking, when cystoscopy reveals the presence of an intravesical tumor, a biopsy is taken for diagnostic purposes, that is to say, to confirm the presence of a tumor and to know its grade and stage. A delay therefore occurs between the time of cystoscopy and receipt of the histopathological diagnosis. Thus, the urine test described herein can be used to determine if the patient suffers from a high-risk tumor almost immediately and thus prioritize their surgery, without having received the final diagnosis of the pathology.

[0074] The test demonstrated excellent results for discriminating stage Ta tumors and carcinoma in situ. These tumors can be difficult to visualize. Thus, the use of the test described herewith can help inform clinicians on a care trajectory adapted to the seriousness of the patient's medical situation.

[0075] Further, also encompassed is the use of the test provided for the reduction in the number of cystoscopies for patients in remission from a bladder tumor. Patients undergo cystoscopy examinations on a recurring basis in the first 24 months and that, thereafter, if they are patients at high risk of recurrence and progression, an annual check-up for life is necessary. The use of the urine test can make it possible to space out control cystoscopies after the first 24 months of remission. Indeed, the urine test canbe used alternately with cystoscopies. Thus, in the presence of a negative urine test result, the patient could wait until a later date for a cystoscopy. Thus, the procedure rate could be reduced by at least 50% after the first 24 months post-resection. For patients that refuse cystoscopy, the provided diagnostic test represent an alternative for these patients.

[0076] The provided combination of four biomarkers allows a sensitivity of 100% for all high-grade bladder tumors, regardless of their stage of development, with a variable specificity of 69-80% depending on the tumor history of the patients. This combination of biomarkers performs well despite the presence of blood in the urine of the patients tested, which makes it interesting as an early screening tool for prioritizing symptomatic patients. Furthermore, since they are stable and abundant proteins, they are easily identifiable, even with less sensitive detection tools, as demonstrated in the validation steps using WB.EXAMPLE I Biomarker Discovery

[0077] All participants included in the Discovery Cohort were enrolled at the Urology Department of the Centre Hospitalier Universitaire de Sherbrooke between 2017 and 2021 following the approval of the research protocol 2018-2619 by the Research Ethics Committee (REC) of the CIUSSS de I’Estrie-CHUS. To ensure informed decisionmaking, each participant received a consent form explaining the research project. Most participants were selected based on their order of arrival for consultation and categorized according to their tumor status. This allowed for the creation of a cohort representative of clinical reality in terms of age, gender, and body mass index (BMI). Patients with renal insufficiency, i.e., those with a glomerular filtration rate (GFR) below 60 ml / min / 1.73m2, and those with macroscopic and microscopic hematuria were also included. Patients with a catheter were excluded.

[0078] As indicated in Table 1 , three major categories of patients were established: (1) without a tumor or a history of bladder cancer at the time of sampling; (2) without a tumor and in remission from bladder cancer at the time of sampling; and (3) with bladder cancer with or without a history of bladder cancer. There were a total of 97 participants, including 33 patients in category 1 , 27 patients in category 2, and 37 patients in category 3 (16 low-grade; 21 high-grade). Medical consultation for patients included in the cohort without bladder tumors was generally motivated by Benign Prostatic Hyperplasia (BPH), Lower Urinary Tract Symptoms (LUTS), the presence of bladder or kidney stones,urethral stenosis, or for kidney or prostate cancer follow-ups. Obviously, the majority of these patients also suffered from some form of hematuria, whether macroscopic or microscopic.

[0079] Each urine sample was collected in standard 50 ml tubes, following the signing of the consent form, and temporarily stored on ice for a maximum of two hours. Urine was collected by self-micturition or during cystoscopy by the treating physician. In the case of internal sampling, special attention was given to the amount of water introduced during the insertion of the cystoscope into the urethra to minimize sample dilution. Once the collection day was completed, all samples were stored in a -80°C freezer until further processing.

[0080] Before being sent to the mass spectrometer, urine samples had to undergo a series of transformation steps to extract and purify the protein content. Each solution prepared contained MS-grade water only.

[0081] Urine samples were thawed at 4°C for a few hours. Once in liquid form, each sample was homogenized using a vortex. For each sample, a duplicate of 3.0 ml of urine was taken from the original tubes and transferred to two 5.0 ml tubes (Protein LoBind Tubes 5.0 ml, Catalog No. 0030108302, Eppendorf©). Thus, 194 samples from 97 participants were prepared (n=2, technical duplicate). Subsequently, all samples underwent two freeze (-80°C) and thaw (25°C) cycles to break the cell membranes and allow the release of their content. Finally, the samples were centrifuged at 4°C for 15 min at 4000 x g in a centrifuge with swinging bucket rotor and sealed caps to prevent biocontamination (SorvallTM ST16R, Thermo Scientific). Once centrifuged, the supernatant was collected and transferred to new 5.0 ml LoBind tubes.

[0082] Proteins in the urine samples were precipitated with 10% V / V trichloroacetic acid (TCA) (Product #T6399, Sigma-Aldrich, St. Louis, USA) overnight at -20°C. After precipitation, the samples were thawed on ice for a few hours and then centrifuged at 20,000 x g at 4°C for 20 min (rotor JLA 25-50, Beckham Coulter). After removing the supernatant, 1 ml of cold acetone was added to the pellet of each sample. They were then centrifuged at 15,000 x g for 15 min at 4°C to remove salts and other impurities. The acetone was finally removed, and the pellets were dried under the hood for 30 min.

[0083] Once dry, the pellets were resuspended with a denaturing solution (8M urea, 10 mM Hepes, pH 7.5), in a variable total volume of 30 to 100 pl depending on the sizeof the pellet. Each sample was then vortexed to ensure homogeneity. The protein content of each sample was subsequently assessed using the colorimetric method with bicinchoninic acid (Pierce BCA Protein Assay Kit, Cat. #23225, Thermo Scientific) and incubated at 37°C for 20 min. The samples were read by spectrophotometry at 562 nm. Protein concentrations were determined using a standard curve.

[0084] A volume corresponding to 10 pg of proteins for each sample was taken and transferred to new 1.0 ml LoBind tubes. The samples were reduced by adding 1 pl of 3.24 mM DTT, heated at 95°C for 2 min, and incubated for 30 min at room temperature. They were alkylated by adding 1 pl of 13.5 mM chloroacetamide before an incubation in the dark for 20 min. A 72 pl solution of 50 mM ammonium carbonate was added as a buffer. Finally, 1 pl of Trypsin Gold 1 pg / pl was added for a final volume of 100 pl. The samples were then incubated at 30°C overnight on a shaker.

[0085] Once digested, the samples were cooled down to 4°C to stop the enzymatic reaction. For each sample, a 100 pl ZipTip© installed on a p200 pipette was used (Pierce C18 Tips, 100 pl, Thermo Scientific) for all purification steps. Firstly, 100 pl of humidification solution (100% acetonitrile) was aspirated and discarded to wet the inside of the mini-column. This step was repeated three times in series. Secondly, 100 pl of equilibration solution (0.1% trifluoroacetic acid (TFA)) was aspirated and discarded into the original tube, three times, to balance the column by positively charging it. Then, the sample containing the digested peptides was aspirated and discarded into the original tube for a total of 7 to 10 cycles. Once the peptides were bound to the C18 column, a washing solution (0.1% TFA) of 100 pl was aspirated and discarded (3 to 5 cycles). Finally, the peptides were eluted with 50% acetonitrile, 0.1% TFA, directly into a new 1.0 ml LoBind tube.

[0086] Once the samples were evaporated, the pellets were resuspended in 25 pl of a 1% formic acid solution. The samples were vortexed for 1 min, then 2 pl was taken to perform the peptide assay at 205 nm using a spectrophotometer (Nanodrop 2000c, Thermo Scientific).

[0087] For each sample to be analyzed, a quantity of 200 ng of peptides was injected in an HPLC (nanoElute, Bruker Daltonics), loaded into a single-flow pre-column constant of 4 pl / min (Acclaim PepMap100 C18 column, 0.3 mm id x 5 mm, Dionex Corporation) then eluted at a flow rate of 500 nL / min on a C18 analytical column (1.9 pm beads size,75 m x 25 cm, PepSep) for a period of 2 hours under a gradient acetonitrile (5-37%) and 0.1% formic acid.

[0088] Once separated and eluted, the peptides were injected into a Tims TOF Pro mass spectrometer whose ion source is produced by electro-nebulization by the CaptiveSpray technology (Bruker Daltonics). The data was acquired by data-dependent auto MS / MS selecting a mass range between 100 and 1700 m / z. The number of scans carried out by the accumulation-sequential fragmentation system parallel (PASEF) was set to 10 (1.27 seconds / cycle) with an exclusion time dynamic of 0.4s, an isolation window dependent on the m / z ratio and an energy of collision of 42.0 eV. The target intensity was set at 20,000 with an intensity threshold minimum of 2500. Each of these steps was carried out twice for each sample (including a blank between each acquisition) in order to be able to evaluate the technical variability of the device and correct it, if necessary, to minimize its impact in the interpretation of the results. Thus, four urine samples from the same patient, i.e. two replicates evaluating the biological variability and two replicates evaluating the technical variability of the LC- MS / MS, were analyzed by LC-MS / MS (n=4).

[0089] The files containing the raw data (.q file), produced following the mass spectrometer analyses, were interpreted, processed and quantified using the software MaxQuant (version 1.6.10) using the UniProtKB database (21-03-2020, 75,776 entries). The parameters used for TIMS-DDA type analysis with MaxQuant were the following: tolerance of 2 missed cleavages; fixed modification including carbamidomethylation of cysteines (effect of chloroacetamide); variable modifications including the oxidation of methionine, the acetylation of the N-terminal regions as well as the carbamylation of lysines and N-terminal regions (effect of urea); Trypsin / P enzyme cleaving after arginines and lysines except before prolines; mass tolerance of 10 ppm for precursor ions and 20 ppm for fragment ions. The threshold identification for the “PSM FRD”, the “Protein FDR” as well as the “Site decoy fraction” were set to 0.01. The number of peptides detected per protein group was set to 1 and a minimum LFQ threshold of 2 was also selected. The “Match between runs” was selected without selecting the “Second peptides search” option.

[0090] Biostatistical analyses were carried out using the R programming language v.3.6.2.

[0091] Identifications of reverse sequences (“reverse”), contaminants (“Potential Contaminants") and proteins having less than 2 unique peptides ("Unique peptides") associated with their identification have been removed. The proteins having been identified but presenting zero values in terms of LFQ intensities for all samples were eliminated. The intensities associated with label-free quantification were transformed into a base 2 logarithm in order to minimize the value differences between the samples and tend towards normality from a statistical point of view. Null values LFQ intensity values were transformed into NA. To determine the batch effect, i.e. the effect of passing through the mass spectrometer on the composition of the samples, a summary analysis of the distribution of intensities was carried out with the R package proBatch 1 .8.0 (Cuklina, 2018, Computational challenges in biomarker discovery from high- throughput proteomic data. ETH Zurich - Doctoral Thesis.). A normalization by quantiles was also applied to the entire LFQ data set. LFQ intensities compiled for each participant correspond to the average of the four replicates. Once the average of the replicates was calculated, the NA values were imputed by the value of the 1st quantile.

[0092] In order to determine the difference in expression of the urinary proteome between the different groups of patients, a base 2 log expression ratio was calculated for each protein detected in urine which passed quality control. This made it possible to generate ratios which were subsequently tested with the non-parametric Mann-Whitney U test to identify proteins whose differential expression, between the two selected conditions, was significant. Indeed, the Mann Whitney U test was used because two groups are being compared at once. Expression ratios whose value was between -1 and 1 were excluded. A minimum threshold of 0.01 has been set for the p value, considering that the search for targets with high sensitivity and specificity are required in the context of biomarker discovery.

[0093] The Metascape analysis tool (Zhou et al., 2019, Nature Communications, 10(V)) was used to generate the general enrichment histograms. Metascape uses the hypergeometric test and Benjamini-Hochberg p-value correction. A higher similarity score to 0.3 with the Kappa test is applied (See Methods: functional enrichment section analysis in the article above for more details). The STRING v.11 .5 analysis tool was used to visualize and export data from dysregulated proteins for each condition to Cytoscape. A confidence threshold of 0.7 was applied for all analyses. The files listing all of the biological processes enriched for each group generated by Markov Clustering Algorithm (MCL) were also taken from the analyses of this platform.

[0094] Cytoscape v.3.8.2 was used to generate the enrichment networks presented in this document. All protein interactions (direct and non-direct) were analyzed via an unsupervised clustering algorithm, the MCL, representing similar interaction partners in the form of subgroups and determine the main associated biological processes. The settings for different MCLs carried out by the software are: granularity of 2.5, 16 iterations and a threshold of 0.7 for as StringDB Score.

[0095] Analyses were carried out with GSEA v.1 .4.0 software (using the Hallmarks gene set database from the Molecular Signature Database (MSigDB) v.7.4. During this type of analysis, all identified proteins are classified in descending order according to the product of the multiplication of the differential expression level (determined by the value of Log2 fold-change) by the p-value determined by the Mann-Whitney U test. Then, an enrichment score (ES) is calculated based on each of the hallmarks listed in the collection. The most enriched proteins with the lowest p-value have a higher positive score. Inversely, the most significant depleted proteins have a higher negative score.

[0096] The proteins chosen for independent validation were selected according to biological criteria and the quality of their identification during their passage on the mass spectrometer. Molecular mass (kDa), number of unique peptides, sequence coverage (free translation of sequence coverage) (%), as well as the MS score established by MS identification are the main parameters that were considered for the selection of antibodies for the next validation step.

[0097] For the participants included in the second cohort, the same experimental design was applied only for the first cohort. There are 104 participants, including 60 patients without bladder cancer at the time of sampling (including 30 patients without a history of bladder tumor and 30 patients in without recurrence) and 44 patients with a positive bladder cancer status (21 low grade, 23 high grade).

[0098] The samples from the second cohort were collected according to those described in the first cohort section.

[0099] Samples stored at -80°C were thawed at 4°C for 2 hours, then 10 ml from each of them was collected and transferred to new tubes. Two cycles of freeze-thaw were carried out to allow cell membrane lysis and release of the cytoplasmic content. The samples were then centrifuged at 4000 x g for 15 min. at 4°C to eliminate insolubledebris and when under TCA precipitation. The protein pellets were resuspended in the denaturing solution (8M Urea, 10 mM Hepes, pH 7.5) then dosed with bicinchoninic acid.

[0100] A quantity of 10 pg of protein per sample was loaded on gel (4-20% Mini- PROTEAN® TGX Stain-FreeTM Protein Gels #4568096) and separated in buffer Trisglycine migration for 30 min (200V - 120mA - 25W). Three urine samples (1 Without tumor, 1 Remission, 1 Cancer) belonging to the discovery cohort were loaded on gels to allow standardization, serve as a comparison compared to LC-MS / MS results and stand as a reference to quantify the protein increase or decrease in samples with cancer compared to cancer-free samples. Transfer to nitrocellulose membrane was carried out for 10 min (2.5V - 1A). The membranes were incubated in a solution of blocking (5% fat- free milk diluted in TBST 1X) for 1 hour at room temperature. Following the washing steps, the primary antibodies were incubated overnight at 4°C on a rotating plate. The list of primary antibodies tested and their dilutions associated is: Fibrinogen p (FGB) 1 :1000 (#sc-271035, Santa Cruz Biotechnology, Inc. USA), Dermatopontin (DPT) 1 :250 (#sc-376863, Santa Cruz Biotechnology, Inc. USA), Histidine-proline rich glycoprotein (HPRG) 1 :250 (#sc-398239, Santa Cruz Biotechnology, Inc. USA), Apolipoprotein A1 (apoA-1) 1 :250 (#sc-376818, Santa Cruz Biotechnology, Inc. USA), Complement Factor H (CFH) 1 :500 (#sc-47685, Santa Cruz Biotechnology, Inc. USA), Amylase (AMY1A, AMY2A, AMY2B) 1 :1000 (#sc-46657, Santa Cruz Biotechnology, Inc. USA), alpha-2- macroglobulin (A2M) 1 :1000 (#sc-390544, Santa Cruz Biotechnology, Inc. USA), Complement C1q subcomponent subunit C (C1 q-C) 1 :1000 (#sc-365301 , Santa Cruz Biotechnology, Inc. USA), Complement C3 (C3) 1 :1000 (#ab133249, Abeam), Napsin A (NAPSA) 1 :5000 (#ab200999, Abeam). The secondary antibodies of the IgG type coupled to the anti-horseradish peroxidase enzyme mouse (#7076S, Cell Signaling Technology) and anti-rabbit (#7074S, Cell Signaling Technology) with a dilution 1 :10000 were used for 1 hour at room temperature then revealed with the UltraScence Western Substrate kit (cat# CCH345-B050MLA, CCH345-B050MLB, FraggaBio).

[0101] The bands were visualized and quantified using ImageLab software (v.1.6.0, BioRad Laboratories Inc. 2020).

[0102] Since no molecule or protein serves as internal standard within the framework of analysis of urine samples, the image obtained following the transfer of the stain-free gel on a nitrocellulose membrane serves as a basis for normalizing the intensity of each band based on its underlying background (total amount of protein transferred). This makes it possible to avoid over- or underestimation of the amount of protein caused bythe transfer step. Thus, the potential bias in the interpretation of the quantification results is prevented.

[0103] A sample belonging to the discovery cohort was chosen so that the quantity of each biomarker it contains can be used to calculate a normalization factor that can allow comparison between the different results obtained on different membranes for the same antibody. Thus, normalization factors were applied to the protein intensity of each group of samples (1 group per membrane) and this, for each antibody tested.

[0104] For antibodies revealing the presence of biomarkers enriched in samples of patients with cancer compared to samples from patients without cancer, sample controls (without history and in remission) were used as a reference point to quantify the change in abundance between the two conditions. Thus, the lowest intensity between the two control samples with cancer-free status without history or in remission was set to 1 thus making it possible to calculate a relative intensity for the other samples from the group. The intensity values presented in the section Results are thus the result of the transformation of raw intensity data into normalized relative intensity values.

[0105] Clinical data used to evaluate the performance of urinary cytology were collected via the Ariane platform from the medical files of the patients included in the discovery and validation cohort, as part of the research protocol submitted and approved by the CER of the CIUSSS de I’Estrie -CHUS. A total of 233 urine cytology results were identified for analysis purposes. They were correlated with those of histopathology in order to evaluate the sensitivity and specificity of this diagnostic method.

[0106] While the present description has been described in connection with specific embodiments thereof, it will be understood that it is capable of further modifications and this application is intended to cover any variations, uses, or adaptations and including such departures from the present disclosure as come within known or customary practice within the art and as may be applied to the essential features hereinbefore set forth, and as follows in the scope of the appended claims.

Claims

WHAT IS CLAIMED IS:1 . A method of diagnosing bladder cancer in a subject comprising: a) providing a urine sample from the subject; and b) measuring level of expressions of at least two protein markers selected from the group consisting of complement factor H (CFH), Fibrinogen p (FGB), alpha-2-macroglobulin (A2M) and pancreatic alpha-amylase (AMY2A).

2. The method of claim 1 , comprising measuring the level of expression of CFH and FGB.

3. The method of claim 1 , comprising measuring the level of expression of CFH and A2M.

4. The method of claim 1 , comprising measuring the level of expression of CFH and AMY2A.

5. The method of claim 1 , comprising measuring the level of expression of FGB and A2M.

6. The method of claim 1 , comprising measuring the level of expression of FGB and AMY2A.

7. The method of claim 1 , comprising measuring the level of expression of A2M and AMY2A.

8. The method of claim 1 , comprising measuring the level of expression of CFH, FGB and A2M.

9. The method of claim 1 , comprising measuring the level of expression of CFH, A2M and AMY2A.

10. The method of claim 1 , comprising measuring the level of expression of FGB, A2M and AMY2A.

11. The method of claim 1 , comprising measuring the level of expression of CFH, FGB, A2M and AMY2A.

12. The method of claim 1 , wherein an increase in expression levels of CFH, FGB or A2M is measured, or a decrease in expression level of AMY2A is measured.

13. The method of any one of claims 1-12, wherein the expression level of at least two protein markers are normalized to a reference value.

14. The method of any one of claims 1-13, wherein the expression level of the at least two protein markers is measured using an antibodies, nanobodies or aptamers.

15. The method of any one of claims 1-14, wherein the expression level of the at least two protein markers is measured by contacting an antibody binding to CFH, FGB, A2M or AMY2A to the urine sample.

16. The method of any one of claims 1-15, wherein the level of expressions of at least two protein markers are measured using an ELISA test, a mass spectrometry analysis or a rapid test.

17. The method of any one of claim 1-16, wherein the urine sample is from a patient with hematuria.

18. The method of any one of claim 1-17, wherein the urine sample is from a patient in remission.

19. The method of any one of claims 1-18, wherein the bladder cancer is a high-grade cancer.

20. A kit for diagnosing bladder cancer in a subject comprising at least two detecting molecules binding to at least two proteins selected from complement factor H (CFH), Fibrinogen p (FGB), alpha-2-macroglobulin (A2M) and pancreatic alpha-amylase (AMY2A), and instruction for use.

21. The kit of claim 20, wherein the detecting molecules are antibodies, nanobodies or aptamers.

22. The kit of claim 20 or 21 , further comprising a filter paper to collect a urine sample from said subject.

23. The kit of any one of claims 20-22, wherein the kit is an ELISA kit, a mass spectrometry kit or a rapid test kit.

Images