Biomarker signatures for prevention and early detection of gastric cancer

Abstract

The present invention provides a method for determining whether a human patient has a pathology that places the patient at risk for and / or requires further medical testing associated with a gastric cancer condition. The present invention relates to an in-vitro method comprising screening a biological sample of blood or plasma where the levels of at least two biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, EGFR, STAT3, and mtDNA are determined, with the proviso that the selected biomarkers do not consist of the set of IL-8 and mtDNA levels. The method may be for assessing the risk of a human patient having non-atrophic gastritis (NAG), or atrophic gastritis / preneoplasia (AG / P), or gastric cancer (GC), or for identifying these risks or the presence of these conditions. In certain embodiments, the method may comprise, as a separate, simultaneous or parallel step, the step of detecting H. pylori infection. The invention also relates to a kit suitable for carrying out such a method, or a set of markers and their use for determining whether a human patient has a pathology that places said patient at risk for a gastric cancer condition and / or requires further medical examination associated therewith, or for prognosing or diagnosing a pre-gastric cancer condition or a gastric cancer condition.

Description

[Technical field]

[0001] The present invention relates to the field of in vitro testing methods based on the search for plasma biomarkers obtained from biological samples, preferably collected from individuals, especially humans, and in particular to methods that can be applied to the prognosis or diagnosis process of gastric or forestomachic cancer conditions, or to the monitoring of patients susceptible to these. Gastric or forestomachic cancer conditions refer to gastric cancer or conditions that can develop into gastric cancer, i.e. conditions that can follow different stages, from asymptomatic to symptomatic, and precede a normal gastric cancer outcome. The present invention is made in the context of a need for early detection of gastric cancer. Thus, the present invention also relates to kits for carrying out the methods of the present invention, sets of markers, and their use. The use of the biomarkers described herein may make it possible to determine whether further clinical investigations should be performed, or to assess the risk, or to finally diagnose the risk or disease, and to assist in the cure of diseases that require early diagnosis in order to avoid poor prognosis. [Background technology]

[0002] introduction Gastric cancer (GC) remains a major public health problem, resulting in approximately 1 million deaths worldwide each year. GC remains the third leading cause of cancer-related deaths and the fourth most commonly diagnosed cancer (1). Although a decline in the incidence of GC has been observed worldwide over the past 30 years, the number of new cases will at best remain constant or begin to increase until 2030 due to population growth and aging. In recent years, new and unusual GC cases have been reported in both low- and high-income countries, primarily in men under the age of 50 (2). These data lead to predictions of an increase in GC incidence, highlighting that this cancer remains a major public health challenge on a global scale. GC is mostly associated with a poor prognosis, with an overall 5-year survival rate of 15%, thereby highlighting the importance of its early detection. GC arises from a multistep process that begins with the development of chronic inflammation, which evolves through preneoplasia (intestinal metaplasia and dysplasia) into cancerous lesions. (3) The major risk factor responsible for 90% of GC cases is Helicobacter pylori infection, which affects half of the world's population.

[0003] Gastric cancer, when diagnosed at advanced stage, has a poor prognosis, but when diagnosed at early stage, it can be a curable disease.However, GC is often asymptomatic or causes only non-specific symptoms in its early stage.By the time severe symptoms occur, cancer has often reached advanced stage and metastasized.Therefore, there is still a need to characterize and validate early GC biomarkers to reduce the morbidity and mortality associated with gastric adenocarcinoma.

[0004] It is important to note that GC can be prevented if precancerous conditions are detected early, and in the best case, before the development of preneoplasia (4). Eradication of H. pylori infection has been proposed to prevent GC. However, eradication of H. pylori infection is not sufficient, since the magnitude of risk reduction depends on the timing of eradication in the preneoplastic process (5). Currently, the diagnosis of GC is limited to gastroscopy, which is usually performed under general anesthesia. It is important to consider that, while this technique involving biopsy is accompanied by an accurate diagnosis when GC is present, in the case of neoplastic lesions, the samples collected to be assayed may be taken from areas of the digestive tract where no lesions are present, so that neoplastic lesions may not always be detected by this invasive technique. This technique is also highly invasive and requires general anesthesia.

[0005] Moreover, unfortunately, screening strategies suitable for large-scale application that would reduce the overall burden of GC diagnosis are also not available. The development of non-invasive methods as blood-based biomarkers is of paramount importance as a powerful contribution not only to diagnostic tools for early detection and prevention of patients at risk of GC, but also to predict disease recurrence / outcome and monitor anticancer treatment, thus improving survival of GC patients.

[0006] The inventors identified biomarker candidates for plasma samples from patients at various stages of the GC cascade using three complementary approaches: i) quantification by enzyme-linked immunosorbent assay (ELISA) of plasma levels of relevant factors selected according to their role in host response to H. pylori infection, inflammation, and tumorigenesis, ii) Proteome Profiler ELISA-based analysis of oncology pathway-related factors (84 proteins), and iii) large-scale screening of plasma proteins by mass spectrometry (MS)-based proteomics. These data led to the presentation of a list of biomarker candidates and signatures that allow to distinguish healthy subjects from patients, especially those at pre-neoplastic and cancer stages. The characterization of these biomarker signatures paves the way for the development of diagnostic tests that allow not only early and easy detection of patients at risk of GC by a simple blood draw, but also individualization of their clinical follow-up.

[0007] Interestingly, the inventors' discoveries have made it possible to determine a set of biomarkers that, when used together (in a so-called "signature"), may provide a powerful and reliable ability to predict or simultaneously predict forestomachic or gastric cancer conditions at various stages (i.e., to distinguish the health and / or disease stage of a patient). Forestomachic or gastric cancer conditions may be identified by the stages defined herein, in particular the AG / P stage and / or the GC stage. Such a tool may be particularly relevant in the monitoring of patients, in particular asymptomatic patients.

[0008] It is noted that, as detailed herein, forestomachic cancer may be part of the carcinogenic process and, in some cases, the presence of forestomachic cancer may precede the appearance of gastric cancer (GC stage) as defined throughout this specification, and therefore, this specification may refer to "gastric cancer condition" as a condition that encompasses "forestomachic cancer condition."

[0009] Therefore, the present invention is based on the experiments described herein and proposes to provide novel means and tools aimed at addressing any one or all of the above-mentioned problems, namely, simple, reliable and efficient biomarkers that allow the determination of whether a human patient has a lesion that puts said patient at risk for a gastric cancer condition (final stage of the gastric carcinogenesis process discussed herein), i.e., in other words, a lesion that puts said patient at risk for developing or having a gastric cancer condition, and ultimately allows the early detection of a gastric cancer condition, including the diagnosis with appropriate accuracy, according to certain embodiments, of the presence of a precursor gastric cancer condition or a gastric cancer condition in a patient. The present invention is particularly aimed at allowing the practitioner to determine the appropriateness of carrying out further medical tests or clinical investigations on the patient in relation to the condition sought. It is a remarkable advantage that the present invention can be carried out on blood or plasma, among the samples taken from the patient. [Prior art documents] [Patent documents]

[0010] [Patent Document 1] WO2015 / 049372 [Non-patent literature]

[0011] [Non-Patent Document 1] Corre and Piazullo, J., Dig. Dis, 2012, 13:2-9. [Non-Patent Document 2] Delacour, H. et al., “LaCourbe ROC (receiver operating characteristic): principals et principales applications en biologie clinique,” ​​Annales de biologie clinique, 2005, 63(2): 145–54 [Non-Patent Document 3] von Wurmb-Schwark et al., 2002, Forensic Science International, 126:34–39 [Non-Patent Document 4] Fernandes et al., 2014, Cancer Epidemiol Biomarkers Re, 23:2430~38 [Non-Patent Document 5] Fan et al., 2009, J Cancer Res Clin Oncol, 135, 983–989 [Non-Patent Document 6] Chatre and Richetti, 2013, J of Cell Sciences, 126:914~926 [Non-Patent Document 7] Graham et al., 1987, Lancet, 1:1174-1177 Summary of the Invention [Means for solving the problem]

[0012] The present invention relates to an in vitro method for determining whether a human patient has a pathology that places said patient at risk for a gastric cancer condition and / or requires further medical examination in connection therewith, comprising screening a biological sample of blood or plasma previously taken from a human patient susceptible to a condition that may develop into a gastric cancer condition, or susceptible to a forestomachic cancer condition, or susceptible to a gastric cancer condition, determining the levels of at least two biomarkers selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA levels, with the proviso that the selected biomarkers do not consist of the set of IL-8 and mtDNA levels; determining the levels of at least two biomarkers selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, EGFR, STAT3, and mtDNA levels, with the proviso that the biomarkers selected do not consist of the set of IL-8 and mtDNA levels; b. comparing the level determined in step a. to a control; and c. If the levels of at least two biomarkers determined and compared in steps a. and b. deviate from their control levels, it is concluded that the human patient has a pathology that places said patient at risk for a gastric cancer condition and / or that further medical examination, in particular clinical investigation, is indicated. The present invention relates to a method comprising the steps of:

[0013] For example, in certain embodiments, when two biomarkers are assayed, if the levels of the two biomarkers, respectively, deviate from their control levels, it is concluded that the human patient has a pathology that places the patient at risk for a gastric cancer condition.

[0014] According to another particular embodiment, when three biomarkers are assayed, if the levels of at least two of said three biomarkers deviate from their respective control levels, it is concluded that the human patient has a pathology that places said patient at risk for a gastric cancer condition.

[0015] According to a particular embodiment, the present invention provides an in vitro method for determining whether a human patient has a pathology that places said patient at risk for a gastric cancer condition and / or requires further medical examinations related thereto, comprising screening a biological sample of blood or plasma previously taken from a human patient susceptible to a condition that is susceptible to developing a gastric cancer condition, or susceptible to a precursor gastric cancer condition, or susceptible to a gastric cancer condition, determining the levels of at least two biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, SELE, MSLN, EGFR, STAT3, and mtDNA, with the proviso that the selected biomarkers do not consist of the set of IL-8 and mtDNA levels; b. comparing the level determined in step a. to a control; and c. If the levels of at least two biomarkers determined and compared in steps a. and b. deviate from their control levels, it is concluded that the human patient has a pathology that places said patient at risk for a gastric cancer condition and / or that further medical examination, in particular clinical investigation, is indicated. The present invention relates to a method comprising the steps of:

[0016] For example, in certain embodiments, when two biomarkers are assayed, if the levels of the two biomarkers, respectively, deviate from their control levels, it is concluded that the human patient has a pathology that places the patient at risk for a gastric cancer condition.

[0017] According to another particular embodiment, when three biomarkers are assayed, if the levels of at least two of said three biomarkers deviate from their respective control levels, it is concluded that the human patient has a pathology that places said patient at risk for a gastric cancer condition.

[0018] The number of biomarkers assayed in the methods disclosed herein can be 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, or 39.

[0019] According to certain embodiments, the number of biomarkers assayed in the methods disclosed herein is between 2 and 6, in particular 2 (provided that the biomarkers selected do not consist of the set of IL-8 and mtDNA levels), 3, 4, 5, or 6.

[0020] According to certain embodiments, the list of biomarkers referred to herein is PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, EGFR, STAT3, and mtDNA levels, with the proviso that the selected biomarkers do not consist of the following set: IL-8 and mtDNA levels.

[0021] The phrase "provided that the selected biomarkers do not consist of the set of IL-8 and mtDNA levels" is intended to state, through the use of the phrase "does not consist of", that when only two biomarkers are selected for level determination (step a.) in the context of the method of the present invention, these markers cannot be IL-8 and mtDNA levels together (i.e., without the presence of at least one additional biomarker as described herein). In the context of the present specification, such markers, IL-8 levels and mtDNA levels, may be further used in combination with one or more additional biomarkers as described herein, but the method described herein does not encompass the use of only these two biomarkers together with IL-8 levels and mtDNA levels. Alternatively, the phrase may be written as "provided that the selected two biomarkers do not consist of the set of IL-8 and mtDNA levels".

[0022] Thus, according to another alternative wording, applicable throughout this application, the condition could be written as "provided that the selected biomarkers do not consist of a strict association with IL-8 levels and mtDNA levels," in which case these specifically listed biomarkers, when used in the methods defined herein or present in the kit or set of markers (or biomarkers) defined herein by "strict association with IL-8 levels and mtDNA levels," are PGK1, CFP, IGFALS, KR It is intended that the present invention will necessarily be associated with at least one additional marker (or biomarker) selected from the following list: T19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-17, TNF alpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN.

[0023] As used herein, SAA1 (SAA2) refers to biomarker SAA1 or biomarker SAA2, or both. These proteins are of the same family and share close sequence homology. Thus, SAA1 and SAA2 may be substituted for each other or may be found synchronically. Thus, in this specification, the term SAA1 (SAA2) may be substituted with "SAA1" or "SAA2" or "SAA1 or SAA2" or "SAA1 and SAA2" in any occurrence, unless the context is irrelevant.

[0024] Please note that, as used herein, the terms "marker" and "biomarker" are used interchangeably (synonyms).

[0025] According to certain embodiments, the in vitro method of the present invention allows, on the contrary, to determine whether a human patient is in a healthy state, i.e., in the present context, whether the human patient does not have a pathology that puts said patient at risk for a gastric cancer condition and / or does not need further medical examinations related thereto, if the biomarkers used allow such a conclusion, as further described in any of the disclosed aspects herein, in particular with regard to the disclosed threshold values ​​for making the decision. Such a method comprises the same steps as described above, in which, in step c., if the levels of at least two biomarkers determined and compared in steps a. and b. deviate from their control levels, it is concluded that the human patient does not have a pathology that puts said patient at risk for a gastric cancer condition and / or does not need further medical examinations, in particular clinical investigations.

[0026] According to a particular embodiment, the present invention provides an in vitro method for determining whether a human patient has a pathology that places said patient at risk for a gastric cancer condition and / or requires further medical examinations related thereto, comprising screening a biological sample of blood or plasma previously taken from a human patient susceptible to a condition that is susceptible to developing a gastric cancer condition, or susceptible to a precursor gastric cancer condition, or susceptible to a gastric cancer condition, determining the level of at least two biomarkers selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-17, TNFalpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN; b. comparing the level determined in step a. to a control; and c. If the levels of at least two biomarkers determined and compared in steps a. and b. deviate from their control levels, it is concluded that the human patient has a pathology that places said patient at risk for a gastric cancer condition and / or requires further medical examination, in particular clinical investigation, associated therewith. The present invention relates to a method comprising the steps of:

[0027] In a more particular embodiment, the method comprises determining the levels of at least two of the biomarkers among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-17, TNFalpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN, where one of the biomarkers selected in step a. is replaced by the level of IL-8 or mtDNA.

[0028] "Lesions that place said patient at risk for gastric cancer" refers to any of the lesions defined above and herein, the severity of which is increasing stepwise according to the stage in the gastric carcinogenesis process that is the subject of this application. For example, lesions may be present as early as the forestomach stage (herein understood as "fore-gastric cancer stage"). For example, lesions at the non-atrophic gastritis (herein abbreviated as NAG) stage may be lesions of chronic inflammation of the gastric mucosa associated with high production of oxidative species. Lesions at the atrophic gastritis (AG) stage may be loss of gastric glands. Lesions at the pre-neoplastic (P) stage may include alterations of gastric epithelial cells, acquiring an enterocyte phenotype or an irregular specific structure of glands. Exemplary lesions at the gastric cancer stage are presented later in this specification.

[0029] In "lesions that render the patient at risk for a gastric cancer condition", "at risk for a gastric cancer condition" is alternatively expressed as "lesions that render the patient at risk for developing or having a gastric cancer condition", and it is intended that the assayed patient, for whom the conditions set in step c. are met, is at risk for developing gastric cancer or is at risk for being in an ongoing gastric carcinogenesis process, including several stages of increasing severity (which may be diagnosed as a precursor gastric cancer condition), even if reversal of the condition may be observed at any stage.

[0030] According to certain embodiments, by "deviation of the levels of at least two biomarkers from their respective control levels", as further detailed herein, is intended a deviation that is statistically significant in the context of a control (which may be a standard for a healthy state or another well-defined state, as long as the change is considered significant according to common practice in the field of determining the significance of the change, in particular statistical significance). Indeed, as will be made clear in this application, the inventors will be able to determine that the variation of the biomarkers of the invention is associated with the presence of a lesion that places the tested patient at risk for a gastric cancer condition and / or requires further medical examination in connection therewith. Thus, in certain aspects, the present invention seeks to first recognize whether it is appropriate for a subsequent search to be sought in relation to the risk of the presence of an ongoing gastric carcinogenic process for the patient being assayed.

[0031] According to certain embodiments, further medical examinations that may be indicated correspond to further clinical investigations, defined as clinical research studies in which the investigator works directly with the patient, in an outpatient or inpatient setting. This definition may include the performance of further in vitro tests, such as performing various blood tests, e.g. a complete blood count (CBC) to check for anemia, but also extends to studies in which human-derived material is obtained via a third party and the investigator does not work directly with the patient. Non-limiting examples of "further clinical investigations" thus include optical examination by gastroscopy, computed tomography (or CT) scanning of the abdomen, biopsies for histological examination, and other exploratory methods aimed at confirming or excluding the presence of a gastric carcinogenic process, the latter of which allows an accurate diagnosis of the lesion, the type of cancer, if present, and / or the stage reached in the carcinogenic process.

[0032] If a gastric lesion is suspected, further clinical investigation may include increasing the number and / or frequency of scheduled optical gastroscopic examinations, which are otherwise performed infrequently.

[0033] The determination of "deviation of biomarker levels with respect to a control level" can be made by any means suitable for this purpose. In particular, to determine "deviation", it can be evaluated whether the level determined for one particular biomarker in step a. is elevated or decreased with respect to a reference value (i.e. cut-off value) measured in a control individual or presented as a control value, such as a control value obtained from pooled values ​​for the control individuals. A specific example of such a determination is shown in Table 2 herein. According to another embodiment, depending on the direction of the change and / or the absolute value of the change, for example expressed as a ratio or in a multiple, and if necessary by comparison with the direction of change known for the analyzed state considered (gastric cancer state or pre-gastric cancer state), a conclusion can be made as to whether the patient has a pathology that places said patient at risk for a gastric cancer state and / or requires further medical examinations related thereto, based on a set of decision rules, for example as shown in Table 7 herein, in step c. above. In fact, as shown herein, experiments such as non-targeted mass spectrometry experiments that result in measurements that are relative values ​​can still provide statistically significant information in a relative form, which is the variation of biomarker levels between disease stages (which can be expressed as a ratio). Conversely, measurements can also be performed using quantitative (targeted) mass spectrometry measurements that result in absolute values ​​that can be compared with each other, rather than the comparison of variations between relative values ​​that is required when using non-targeted mass spectrometry experiments, examples of which are shown in the experimental section of this specification. The present invention can be performed in any case where deviations for assayed parameters, as defined herein, are determined against absolute reference values ​​or using a comparison between the variations of pooled values ​​between distinct disease stages or health conditions. Such variations can be expressed as a ratio. Examples are provided in the experimental section of this specification.

[0034] According to a particular embodiment, step a. as defined above comprises determining the levels of at least three biomarkers, including two biomarkers, IL-8 protein level and mtDNA level, further combined with one or more biomarkers selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-17, TNF alpha, USF1, USF2, SELE and MSLN. determining the levels of at least three biomarkers, optionally including two biomarkers being IL-8 protein levels and mtDNA levels, further combined with one or more biomarkers selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-17, TNF alpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN.

[0035] According to a particular embodiment, step a. comprises: - levels of PGK1 further combined with one or more biomarkers selected from among levels of CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of CFP further combined with one or more biomarkers selected from the levels of PGK1, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of IGFALS further combined with one or more biomarkers selected from among the levels of PGK1, CFP, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of KRT19 further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of SPRR1A further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of CPA4, further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of CA2, further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of SERPINA5 further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of MAN2A1 further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of KIF20B, further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of SPEN, further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of JUP, further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of KRT6C further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of CDSN further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of KPRP further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of F13A1 further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtD the level of SAA1 (SAA2), further combined with one or more biomarkers selected from among the level of NA (in the context of this paragraph, SAA1 (SAA2) means "SAA1" or "SAA2" (but in this case, other biomarkers, SAA1 or SAA2, respectively, that are omitted from the list, may also be added, or in the context of this paragraph, SAA1 (SAA2) may mean "SAA1 and SAA2"); - the level of LBP, further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of DSP, further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of KRT2 further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of KRT14 further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of ARG1 further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of S100A12 further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of ATAD3B, further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of MAN1A1 further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of HAL further combined with one or more biomarkers selected from the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of DCD further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of C7 further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of HP further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of LEP further combined with one or more biomarkers selected from the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA; - the level of IL-8, further combined with one or more biomarkers selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-17, TNF alpha, USF1, USF2, SELE, and MSLN; - the level of IL-17, further combined with one or more biomarkers selected from the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8TNF-alpha, USF1, USF2, SELE, MSLN, and mtDNA; - levels of TNF-alpha, further combined with one or more biomarkers selected from among levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, USF1, USF2, SELE, MSLN, and mtDNA; - the level of USF1 further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF2, SELE, MSLN, and mtDNA; - the level of USF2 further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, SELE, MSLN, and mtDNA; - the level of SELE, further combined with one or more biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, MSLN, and mtDNA; - the level of MSLN further combined with one or more biomarkers selected from the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, and mtDNA; - the level of mtDNA further combined with one or more biomarkers selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-17, TNF alpha, USF1, USF2, SELE, and MSLN; - the level of IL-8 and mtDNA levels further combined with one or more biomarkers selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-17, TNF alpha, USF1, USF2, MSLN, and SELE; The method may include determining:

[0036] According to certain embodiments, the biomarkers EGFR and STAT3 are added to the one or more biomarkers listed above, with the proviso that if IL-8 levels and mtDNA levels are determined in combination, the level of at least one additional biomarker is also determined.

[0037] According to certain embodiments, "one or more" means a total of three, four, five, or six biomarkers, i.e., "multiple" means two, three, four, five, or six biomarkers.

[0038] In certain aspects, the present invention does not encompass embodiments in which the selected biomarkers consist of a strict association with IL-8 levels and mtDNA levels.

[0039] According to certain embodiments, the biomarkers assayed are as shown in Table 4, Table 5, Table 6, Table 7, Table 8, Table 9, Table 10, Table 11, or Table 12, and / or the combinations depicted in any of Figures 9, 11, 13, or 15.

[0040] According to certain embodiments, the biomarkers assayed include at least one biomarker selected from S100A12, KIF20B, ARG1, DSP1, or HAL. S100A12 has been shown to be relevant in assessing gastric cancer risk, and KIF20B, ARG1, DSP1, and HAL have been shown to be relevant in assessing AG / P stage. When selected, one or more of these markers may be associated with any combination of biomarkers described herein.

[0041] According to a particular embodiment, step a. of the method of the invention consists of determining the levels of at least two biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, SELE, MSLN and mtDNA, optionally together with at least one further biomarker selected from among EGFR and STAT3. In a particular embodiment, step a. of the method of the invention consists of determining the levels of at least three biomarkers, where at least two biomarkers are selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, SELE, MSLN, and mtDNA, and optionally the third biomarker is selected from among EGFR and STAT3.

[0042] According to a particular embodiment, step a. of the method of the invention consists of determining the levels of at least three biomarkers, including two biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, SELE, MSLN and mtDNA, and at least one further biomarker selected from among EGFR and STAT3.

[0043] According to a particular embodiment, step a. of the method of the invention consists of determining the levels of at least three biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, SELE, MSLN, EGFR, STAT3, and mtDNA.

[0044] According to a particular embodiment, step a. of the method of the invention consists of determining the levels of at least two, preferably between two and six, biomarkers selected from among IGFALS, KRT19, CPA4, CA2, MAN2A1, KIF20B, JUP, F13A1, LBP, KRT14, ARG1, S100A12, ATAD3B, DCD, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, SELE, MSLN, EGFR, and STAT3.

[0045] According to a particular embodiment, step a. of the method of the invention consists of determining the levels of at least two, preferably between two and six biomarkers (determined via ELISA experiments) selected from among IGFALS, KRT19, CA2, MAN2A1, KIF20B, JUP, LBP, ARG1, S100A12, ATAD3B, DCD, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, SELE, MSLN, EGFR, and STAT3, as shown, for example, in FIG. 13 and Table 12.

[0046] According to a particular embodiment, step a. of the method of the invention consists of determining the levels of at least two, preferably between two and six, biomarkers selected from among IGFALS, KRT19, CA2, MAN2A1, JUP, ARG1, S100A12, HP, LEP, IL-17, TNFalpha, USF2, SELE, MSLN, and EGFR.

[0047] According to certain embodiments, the deviation of the pooled values ​​of the levels of at least two biomarkers between distinct disease stages / health states, as determined using, for example, a comparison between Table 7 (Table 9), from their respective corresponding reference levels, indicates the variation in protein levels between distinct disease stages / health states, expressed as a ratio (which can be converted to a fold change; see description). For the protein DCD, the average log2 ratio change between the healthy and pre-neoplastic patient pools was observed to be 0.76, which is 2.54×10 -2 It can be seen that the log2 mean change in ratio between healthy and pre-neoplastic patients is found to be -0.64, which means that the change in this ratio (direction of change and approximate magnitude of change) is considered significant to conclude that there is a deviation in the levels of the protein DCD relative to healthy patients. Similarly, for the protein LEP, the mean log2 change in ratio between the healthy and pre-neoplastic patient pools is observed to be -0.64, which is 5.91 x 10 -2It can be seen that the p-value of DCD is found to be associated with a p-value of 0.01 to 0.01. This means that a change in this ratio (direction of change and approximate magnitude of change) is considered significant to conclude that there is a deviation in the level of the protein LEP with respect to healthy patients. In a situation where both the protein levels of DCD and LEP are assayed and both parameters show a deviation in the direction of change and approximate magnitude of change as indicated above with respect to the average value of the pool of healthy patients collected as a reference value, it may be said that the levels of the two biomarkers deviate from their control levels, respectively, and that the patient from whom the sample was taken may be concluded to be at risk for a gastric cancer condition, in the sense that the observed variation indicates a risk of gastric preneoplasia. This example is presented to illustrate the teaching that can be drawn from Table 7 (Table 9): that such an inference can be made for all the biomarkers depicted and all the conditions assayed. Table 8 also presents as an exemplary reference for those skilled in the art that values ​​for the plasma levels of the biomarkers identified in the signatures predictive of pre-neoplasia and GC can be readily compared to reference values ​​that can be collected from samples of individuals determined to be healthy for the test condition.

[0048] According to certain embodiments, P-values, which are applicable in any part of this description and which are in line with conventional practice in the field of medical statistics, are considered to define a statistically significant test when the value is below 0.05 (P-value<0.05), and in some cases, equal to or less than 0.05, as would be appreciated by one of skill in the art.

[0049] According to certain embodiments, each assayed biomarker may be included in a decision rule that may be associated with an AUC value that indicates significant predictive power when greater than 0.5 and indicates perfect prediction when at most equal to 1 for the assayed condition or among several conditions. According to certain embodiments, the assayed biomarkers result in a test that may be determined to be associated with an AUC value of at least 0.5, preferably at least 0.6, or at least 0.65, or at least 0.7, or at least 0.75, or at least 0.8, or at least 0.85, or at least 0.9, or at least 0.95, and in particular an AUC value of 1, for the assayed condition. The skilled artisan knows the means to calculate AUC values, and complete guidance is provided in the present description. For some marker combinations, AUC values ​​can also be calculated using first, estimation of a multinomial logistic regression model (see experimental section for exemplary, non-limiting protocol), followed by the determination of a decision rule based on the estimated multinomial logistic regression model (see multinomial logistic regression model in this description).In addition, a statistical analysis is used by residual deviation criterion, which allows to determine how well the model fits all disease stages.See experimental section for exemplary guidelines.

[0050] According to certain embodiments, in particular when the in vitro method described herein is accompanied by specificity and sensitivity parameters that can be attributed to the test according to one or more cut-off values ​​described herein, the in vitro method described herein can be said to be a method for prognosing or diagnosing a gastric cancer condition or a forestomach cancer condition, but not necessarily. Herein, examples of possible determination outcomes associated with specificity and sensitivity parameters are presented (for example, the data summarized in Table 2 for single parameters or double measurements, or the ELISA results shown in Table 4 or Table 5 for double or triple measurements. Specificity and sensitivity may vary depending on the cut-off value used for determination and can be defined in an optimized manner using ROC curves).

[0051] "Human patients susceptible to developing a gastric cancer condition or a pre-gastric cancer condition" may include patients who present risk factors for the development of gastric cancer, for example due to physical clues, family history, or complaints that indicate to the practitioner that a cause of gastric cancer may be present or may become present. It also includes patients with gastroesophageal reflux, patients with chronic gastric pain, as well as H. pylori seropositive or H. pylori seronegative subjects, who have previously been eradicated from H. pylori infection. This definition also includes individuals who have a condition susceptible to developing a gastric cancer condition or who have been informed of a gastric cancer condition, whether under treatment or not, and who should be monitored. Particular patient groups that are eligible for the implementation of the method of the present invention are patients with chronic inflammation associated with gastritis, or patients with gastroesophageal reflux, patients with chronic gastric pain, as well as H. pylori seropositive subjects. As mentioned above, "patients susceptible to suffering from a gastric cancer condition or to suffering from a condition that is susceptible to developing into a gastric cancer condition" may also be H. pylori negative at the time of collection and / or testing due to successful eradication of the infection. The above definition also includes patients already diagnosed with chronic atrophic gastritis or other gastric pathologies that require clinical follow-up. However, since the present invention is aimed at early detection of gastric cancer, such as molecular disease stage, based on the determination of biomarker levels in blood or plasma, easily and without warning clinical signs, this definition also includes patients who are completely asymptomatic in light of any clinical clues commonly associated with gastric cancer.

[0052] By "gastric cancer condition" is intended gastric cancer as conventionally diagnosed according to medical practice. It includes both cardia cancer (upper gastric cancer) and non-cardia cancer (mid- and distal gastric cancer). "Gastric cancer" may encompass diffuse gastric adenocarcinoma, intestinal gastric adenocarcinoma, and MALT lymphoma. MALT lymphoma (or MALToma) is a form of lymphoma related to the gastric mucosa-associated lymphoid tissue (MALT). MALT lymphoma is frequently (but not in all cases) associated with chronic inflammation resulting from the presence of H. pylori or is linked to the presence of H. pylori. Gastric adenocarcinoma is a malignant epithelial tumor derived from the glandular epithelium of the gastric mucosa. Gastric adenocarcinoma represents the majority of GCs, i.e., more than 90% of diagnosed GCs are adenocarcinomas. Two types of gastric adenocarcinoma: intestinal type or diffuse type, are based on histological distinction. Different stages can be associated with GC using known classification systems, such as the staging system according to the "TNM Classification of Malignant Tumours", which describe the spread of cancer in patients. Using this type of classification, for example, stages 0, I, II, III, or IV can be distinguished. In stage 0, gastric cancer is limited to the lining of the gastric mucosa and may be treatable by surgery, without the need for chemotherapy or radiation treatment, if found very early. In stages I and II, the disease invades the deep layers of the gastric mucosa and may be treated by surgery, possibly accompanied by chemotherapy and / or radiation treatment. In stage III, the disease may have invaded other nearby tissues and distant lymph nodes. In stage IV, the disease has spread to nearby tissues and more distant lymph nodes or has metastasized to other organs. In this description, gastric cancer is generally abbreviated as "GC", unless otherwise indicated or the context dictates otherwise.

[0053] Intestinal-type GC, often but not always induced by H. pylori infection, develops through a series of precursor lesions. Thus, by "pre-gastric cancer state" is intended a range of events from non-atrophic gastritis (abbreviated herein as NAG) or atrophic gastritis (AG) to pre-neoplasia (P), which correspond to chronic inflammation of the gastric mucosa associated with high production of oxidative species, as described in Corre and Piazulo, J., Dig. Dis, 2012, 13:2-9. AG is the first recognizable stage of the pre-cancerous cascade, corresponding to the loss of gastric glands. Pre-neoplasia encompasses intestinal metaplasia (IM) and dysplasia, prior to the gastric cancer stage. IM and dysplasia are recognized as pre-neoplastic lesions. IM corresponds to the alteration of gastric epithelial cells, acquiring an enterocyte phenotype. IM is a state that predisposes to malignancy. Dysplasia is also referred to as a non-invasive neoplasm involving irregular specific structures of secretory glands.

[0054] As mentioned above, since H. pylori infection is a major risk factor for GC, a "condition predisposed to developing gastric cancer" includes, for example, H. pylori infection in a patient.

[0055] According to a particular embodiment, the present invention to this effect seeks to more accurately assess the risk according to the present description, in particular the risk that a human patient has of developing or having a gastric cancer condition, immediately prior to gastric cancer, i.e. in the stages or conditions from atrophic gastritis (AG) to preneoplastic (P), also referred to herein as AG / P, by reference to a cohort of test patients, and in particular to seek prognostic or diagnostic methods for a pregastric cancer condition affecting a test individual.

[0056] Thus, the present invention is also a method for assessing the risk of a human patient having atrophic gastritis / preneoplasia (AG / P), and in particular relates to a method for prognosing or diagnosing an atrophic gastritis / preneoplasia (AG / P) condition in a test patient.

[0057] In a specific embodiment, the invention is a method for prognosing or diagnosing an atrophic gastritis / preneoplastic (AG / P) condition in a test patient, if the outcome permits.

[0058] In a particular embodiment, the present invention is a method for assessing the risk of a human patient having non-atrophic gastritis (NAG), or atrophic gastritis / pre-neoplasia (AG / P), or gastric cancer (GC), if the outcome permits; in particular, the present invention relates to a method for identifying the presence of non-atrophic gastritis (NAG), atrophic gastritis / pre-neoplasia (AG / P), gastric cancer (GC), or a healthy condition in a test patient, and in particular, a method for simultaneously prognosing or diagnosing one of these conditions or the other, among others.

[0059] The methods described herein are based on the detection or monitoring of a biological parameter for a patient and / or allow the provision of information about the health status of such a patient. Where a sensitivity / sensitivity value can be associated with the test performed, the methods of the invention can be defined as methods allowing a diagnosis of the desired condition.

[0060] Thus, the search method described herein makes it possible to determine, according to a non-statistical definition, at least the risk of the presence of a forestomach cancer condition in a patient whose sample is assayed according to the method described herein, or to determine the likelihood of this occurrence or risk. According to a different embodiment, said determination reaches a prognosis or diagnosis of a forestomach cancer condition in a patient, if the involvement does not rely on further tests or clinical investigations, and the comparison of the collected values ​​with the control values ​​of involvement makes it possible to conclude, as a direct result, about the presence or absence of a forestomach cancer condition. According to common practice in the art, the sensitivity / sensitivity value associated with such a determination corresponds to the selection of a threshold value, which is optimized and maintained according to common knowledge in the art, in particular by the use of ROC curves, as discussed later in this specification.

[0061] According to a particular embodiment, the method of the invention is carried out on a blood sample taken from a human patient suffering from a gastric cancer condition or a pre-gastric cancer condition, or suffering from a condition predisposing to the development of a gastric cancer condition.

[0062] In a particular embodiment, the invention is based on the measurement of biomarkers in plasma. Thus, the biological sample may be a blood, plasma or serum sample.

[0063] In case protein biomarkers are required, the invention is advantageously carried out on the basis of a biological plasma sample already obtained from a human subject. This sample can be isolated (collected, taken) from an individual susceptible to suffering from a gastric cancer condition, or susceptible to suffering from a pre-gastric cancer condition or a condition susceptible to developing into a gastric cancer condition, as defined above. The individual may or may not have already been diagnosed with gastric cancer or a pathology that may lead to gastric cancer (pre-neoplastic or pre-gastric cancer condition), and these individuals may optionally be under treatment, such as surgery and / or chemotherapy and / or radiation treatment.

[0064] A plasma sample is the liquid portion of a blood sample that carries the cells and proteins that blood contains. It is noted that serum is plasma without the clotting factors. According to certain embodiments, where relevant, the invention can also be carried out on samples that are serum samples.

[0065] Although the present invention focuses on measuring protein levels found in an individual's plasma, according to certain embodiments, mitochondrial DNA (herein abbreviated as "mtDNA") levels may also be measured. This is conveniently done by testing for mtDNA in white blood cells found in the blood (albeit in a plasma sample that does not contain cells). However, since mtDNA is found in circulating blood, the level of mtDNA considered may also be referred to as a "plasma" biomarker. According to a preferred embodiment in which the level of mtDNA is measured, the level of mtDNA is determined by testing for circulating mtDNA, in particular by testing for mtDNA in white blood cells.

[0066] According to a particular embodiment, the biological sample taken from the test individual is a blood sample. According to a more particular embodiment, such a blood sample is pre-processed to separate white blood cells from which total DNA is prepared and purified. Measurement of the level of mtDNA can then be performed on the collected total DNA preparation of white blood cells, if necessary, in parallel with the measurement of other plasma biomarkers found in the plasma of the same sample containing the separated white blood cells.

[0067] The invention is more particularly based on the measurement of the level of at least two biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, SELE, EGFR, STAT3 and MSLN proteins, as defined in step a. above, optionally with the conditions disclosed in the present description and in any embodiment described.

[0068] Since these markers are proteins, they may be conveniently measured by enzyme-linked immunosorbent assay (ELISA) tests or, alternatively, by mass spectrometry, based on samples obtained from the test individual. According to another embodiment, these markers may also be measured via Luminex experiments, following guidelines available to the skilled artisan in the literature.

[0069] Some of these proteins have been determined to be suitable targets for the purposes of the present invention through large-scale screening of plasma proteins by mass spectrometry (MS)-based proteomics. In a particular embodiment, the protein combinations shown in Table 6 or FIG. 11 herein are combinations of at least two plasma biomarkers of interest, in particular combinations of three plasma biomarkers of interest, associated with good predictive AUC values, determined using the protocol shown in the experimental section, for discriminating the risk of being in the NAG stage, or the AG / P stage, or the GC stage, or the presence of the NAG, AG / P, or GC state in a test individual. The combinations of plasma biomarkers are provided in the context of so-called "ELISA" biomarkers, including mtDNA levels, in FIG. 9. Some biomarkers have been confirmed via ELISA, and promising biomarkers are summarized herein (see Tables 8-12).

[0070] Protein levels in plasma can be measured using methods known in the art, such as, by way of non-limiting example, via enzyme-linked immunosorbent assay (ELISA) (Engvall E and Perlman P, 1971, Immunochemistry, 8:871-874). Examples of plasma proteins are provided in the experimental section herein (for example, plasma levels of selected different biomarker candidates may be assessed via commercially available ELISA assays from Duo Set RD system or MyBioSource: Interleukin 8 (IL-8) (Reference No. DY208); Interleukin 17 (IL-17) (Reference No. DY317); Tumor necrosis factor alpha (TNF-α) (Reference No. DY210); Mesothelin (MSLN) (Reference No. DY3265); E-selectin (SELE) (Reference No. DY724); Haptoglobin (HP) (Reference No. DY8465); Leptin (LEP) (Reference No. DY398); and Upstream Stimulatory Factor 1 and Upstream Stimulatory Factor 2 (USF1; Reference No. MBS9342772, and USF2; Reference No. MBS9321077; MyBioSource)). One of skill in the art can readily determine and find the appropriate ELISA reagents to measure the desired protein levels in a sample.

[0071] Luminex assay is a bead-based simplex or multiplex immunoassay system in a microplate format. In multiplex form, Luminex assay detects many targets, for example up to 500 targets, simultaneously in a single sample, while the agents that allow the capture of the targets coupled to the beads can be of different types, i.e., proteins, including antibodies, ligands, and nucleic acids, specific for the desired targets (within a single assay). The beads that can be used in Luminex assay can have different spectral addresses (so-called "color codes"), for example, intrinsically labeled beads with two fluorophores with different ratios. The beads can also be magnetic or non-magnetic beads. In Luminex assay, the sample to be analyzed is added to a mixture of color-coded magnetic or non-magnetic beads that are pre-coated or coated with analyte-specific capture agents, such as antibodies. For example, if the agent is an antibody, a biotinylated detection antibody specific to the analyte of interest is added, and the antibody forms an antibody-antigen sandwich after binding to the analyte of interest. Phycoerythrin (PE)-conjugated streptavidin can also be added to bind to the biotinylated detection antibody. The beads are then read on a dual-laser flow-based detection instrument, i.e., one laser classifies the beads and determines the analyte detected. The second laser determines the magnitude of the PE-derived signal, which is directly proportional to the amount of analyte bound. Magnetic beads are also used to hold the magnetic beads in the monolayer, while two spectrally distinct lights emitted by, for example, light-emitting diodes (LEDs) illuminate the beads. One light identifies the analyte detected, and the second light determines the magnitude of the PE-derived signal. Luminex allows high-throughput experiments and is powerful when exploring changes in the concentration of multiple targets as stated above. Kits for performing Luminex experiments may include capture beads that are conjugated or can be conjugated to the capture agent used, such as a capture antibody.The beads may be color-coded beads, and / or magnetic or non-magnetic beads, and / or carboxylated beads. If the beads are carboxylated beads, the kit may include an amine coupling kit for attaching a capture agent, such as an antibody, to the beads. The kit may include a biotinylated antibody as a detection (secondary) antibody, and / or may include phycoerythrin (PE)-conjugated streptavidin to reveal the biotinylated antibody. Such assays may be carried out according to the instructions and guidelines provided by the manufacturers, such as BioRad, ThermoFischerScientific, Luminex, or R&D Systems, to name a few, and according to conventional knowledge in the art, as detailed in the literature.

[0072] The manner of "measuring the level" of a protein biomarker may depend on the technique used to measure it. If an ELISA test is used, the level of the protein biomarker may be conveniently given as a concentration (quantitative value), using the common practice in the art. If mass spectrometry is used, the level of the protein biomarker in the assayed plasma sample may be conveniently converted into an arbitrary value, whether representing an absolute value (targeted mass spectrometry experiments) or a relative value that allows to determine the variation in the context of the groups used for comparison (non-targeted mass spectrometry experiments). In particular, to obtain an "arbitrary value" based on actual physical measurements, it is possible to use a proteomic analysis by means of an array consisting of capture antibodies present on a nitrocellulose membrane, incubated with a plasma sample, mixed with a cocktail of biotinylated detection antibodies and revealed by streptavidin-horseradish peroxidase, as shown in the experimental section of this specification. As a result, the value obtained for the presence of a protein in the assay sample can be evaluated as an intensity value, and this latter value can be further used for the statistical treatment of the information. Comparison with intensity values ​​obtained for separate groups of individuals, including healthy individuals, is readily and well within the purview of one skilled in the art. According to another embodiment, "measurement of levels" of protein biomarkers may use quantitative polymerase chain reaction (q-PCR), for example via TaqMan protein assays, which allow measurement of both protein levels and mtDNA levels from plasma and leukocytes collected simultaneously from patients, based on the same technique. TaqMan protein assays routinely allow quantification of sample proteins using real-time PCR and antibodies.

[0073] Once the measurement of the levels of at least two biomarkers has been made according to step a., according to the method of the invention, and in order to be able to assess the risk, in particular from prognosing or diagnosing a gastric cancer condition or a pre-gastric cancer condition, according to the present description, the levels determined and obtained in step a. are "compared" to a control, as already discussed. A synonym of "control" can typically be a so-called "reference" value obtained for the same assay, but for an individual or a pool of individuals known to be healthy or having a specifically determined condition for the test condition (see Table 7). However, it is worth mentioning that the skilled person can easily determine such a control value, if necessary, by reviewing the literature in the art for the relevant test condition, and can also adjust and optimize such a "control" value for that purpose, depending precisely on the test cohort, the test patients, using, inter alia, the well-known ROC curve method. The experimental section provides further guidance in this regard. In certain embodiments of the methods of the invention, the control can be an internal control, for example, where a patient's health is monitored by testing biological samples at multiple time points over time.

[0074] According to certain embodiments, the "control" value can also be defined as a "threshold" that allows a decision to be made, i.e., a value that is considered to be "normal". Such a threshold is generally determined for subjects that are determined to be healthy. An example is presented in Table 2. The normal threshold determined for a healthy subject can be a value found in the literature, i.e., a value known to represent a healthy situation, or a value found by assaying one biological sample from a healthy subject, or alternatively, a value found by assaying several biological samples from several separate healthy subjects, and then the resulting normal threshold is determined as the mathematical mean value of the level values ​​of all the biological samples of the healthy subjects that are assayed, or alternatively, a value found by assaying a pool of biological samples from several separate healthy subjects. By "healthy subject" is intended a subject that does not have symptoms of gastric disorders, or a patient referred to gastroscopy with gastric biopsy that corresponds to a normal phenotype. According to a particular embodiment, a "control" value is a value found in a healthy individual (or a group thereof; see above), who has been determined to be healthy by a person skilled in the art in the field of the invention and for this purpose by generally accepted standards. Control groups of healthy volunteers with negative H. pylori serology and / or asymptomatic individuals who are not suspected of the disease or condition in question in the present invention are encompassed within such a definition. Conversely and according to another embodiment, the control group for defining the value is set for individuals considered to be healthy in the medical field by the best known standards.

[0075] Examples of thresholds can be found in the experimental section of this specification, for example, for factors IL-8, IL-17, and TNF alpha, with the decision rule in ng / mL, along with the sensitivity and specificity of the particular values ​​for the decision made. Examples are also presented for factors USF1 and USF2, with further specification of the AUC value determined by ROC curve analysis, the determination of ROC curves being known by those skilled in the art. Figures 3 and 4 also present examples of the determination of ROC curves. ROC curves represent the derivation of true positive rate (TPR) as a function of false positive rate (FPR), obtained by processing data obtained by experiments carried out based on decision rules such as "if the amount of biomarker is above a cut-off value x (or, depending on the configuration, below x, or above or below x, the patient is H or NAG or AG / P or GC (selected as appropriate)". This is made possible by the multiplicity of experiments carried out, which allows the skilled person to easily determine appropriate "control" values, "normal" values, "cut-off" values, "threshold values" as necessary for the purpose of making a decision. Furthermore, the parameter AUC, i.e. the parameter "area under the curve", is a useful way to generalize the overall diagnostic accuracy of a decision rule. AUC takes values ​​between 0 and 1, where a value of 0 indicates a completely inaccurate decision rule and a value of 1 reflects a completely accurate decision rule. If the AUC is below 0.5, it means that the decision rule is useless since it does not outperform a random decision. The AUC can be calculated using rules known by the skilled person and also known in the literature (Delacour, H. et al., "LaCourbe ROC (receiver operating characteristic): principles et principales applications en biologie clinique", Annales de biologie clinique, 2005, 63(2):145-54).

[0076] Thus, one skilled in the art can easily determine, in all circumstances, the "optimal cut-off value", i.e., the control value in the context of the claimed invention, if relevant, using a compromise between the TPR and the FPR.

[0077] It is noted that the skilled artisan may also envisage simply analyzing the direction of change in biomarker levels measured in an individual by reference to a pool of values ​​(see Table 7), if required, by defining a "control" value, examining the "fold" (or "ratio") of the measured or experimental value relative to a "normal" situation (or another "known" state for a reference group), or by simply examining whether the direction of change is the same or different relative to a known change. A "fold" number refers to an expression of change, in particular a number that describes how much a given amount changes from a normal value to a test value, with the normal value being, in particular, the "normal threshold value" and the test value being the value for the test sample. For example, a normal value of 30 and a test value of 60 corresponds to a 2-fold change, or in common terms, a 2-fold increase. Conversely, a normal value of 60 and a test value of 30 corresponds to a 0.5-fold change, or in common parlance, a 0.5-fold increase, also referred to as a "minus" 2-fold increase (denoted in negative terms by a "minus" sign before the number). Thus, the fold change corresponds to the ratio of the test value to the normal value. In other words, the fold change results from the determination of the ratio of the test value to the normal value.

[0078] Table 7 shows the variability profile between assay samples according to several comparison schemes. When making a comparison with healthy samples, the changes depicted in this table and expressed as log2 (ratio of change) in the context of the analyzed parameters may be used with the same convention as the reference to the "control value", the ratio / change carrying the same information as that provided by the comparison of the specific value with the control value. The log2 value is calculated according to the formula: xThe fold change values ​​can be converted into corresponding fold change values ​​using the fold change function. Definition of several parameter combinations allows, at a first level, to distinguish (for example) healthy patients from diseased patients, and at another level, to examine whether diseased patients relate to a particular patient type along the carcinogenic cascade. All point-to-point variations shown in Table 7 (Table 9) are part of the present invention and can be referred to the direction of change they indicate and the order of magnitude of fold / ratio they represent. These variations can provide a table to identify whether patients relate to a particular group along the carcinogenic cascade.

[0079] According to a particular embodiment, the method of the present invention comprises: - protein levels of PGK1 and CFP, or - protein levels of KIF20B and SPEN, or - protein levels of JUP and KRT6C, or - protein levels of JUP and CDSN, or - protein levels of JUP and KPRP, or - protein levels of F13A1 and SAA1 (SAA2), or - protein levels of KRT19 and LBP, or - protein levels of DSP and KPRP, or - protein levels of DSP and CDSN, or - Protein levels of KRT2 and CDSN The method is carried out by assaying for a biomarker that is

[0080] These biomarkers have been shown to be relevant, relative to each other, with excellent AUC values ​​for the assessment of risk according to the present description, or for the prognosis and diagnosis of the presence of an AG / P condition in a subject individual, as depicted in Table 6 herein.

[0081] According to another embodiment, and in keeping with the discussion of FIG. 12, the methods of the present invention involve assayed biomarkers including protein levels of at least KIF20B and SPEN, alone or in further combination with at least another biomarker selected from among protein levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, and C7, and in particular in this case, assayed biomarkers include protein levels of at least KIF20B and SPEN, in further combination with at least another protein biomarker selected from among protein levels of KRT19, ARG1, DSP1, and HAL.

[0082] Indeed, the accompanying experimental results show that the combination of KIF20B and SPEN provides good prediction of AG / P or GC.

[0083] According to a particular embodiment, the method of the present invention comprises: - protein levels of KIF20B, SPEN, and KRT19, or - protein levels of KIF20B, SPEN, and ARG1, or - protein levels of KIF20B, SPEN, and DSP, or - protein levels of KIF20B, SPEN, and HAL, or - protein levels of KIF20B, SPEN, and C7, or - protein levels of KIF20B, SPEN, and JUP, or - protein levels of KIF20B, SPEN, and CPA4, or - protein levels of KIF20B, SPEN, and SPRR1A, or - protein levels of KIF20B, SPEN, and MAN2A1, or - Protein levels of KIF20B, SPEN, and KRT14 The method is carried out by assaying for a biomarker that is

[0084] These biomarkers are shown to be relevant in the association of at least three biomarkers for risk assessment according to the present description, in particular for prognosis and diagnosis of the presence of AG / P status in a test individual, with excellent AUC values, as depicted in Table 6 herein. Furthermore, and as an added advantage, these biomarkers are accompanied by optimal residual deviance for simultaneously predicting each of all pathologies, NAG status, AG / P status, and GC status, as shown in FIG. 11B.

[0085] According to another embodiment, and in line with discussion of FIG. 12, the present invention relates to a method according to any embodiment described herein, further comprising determining the level of at least one additional biomarker selected from among the levels of mtDNA, MSLN, HP, SELE, and TNF-alpha protein levels.

[0086] These markers have indeed been shown to be suitable for prognosing or diagnosing AG / P conditions in patients.

[0087] With regard to mtDNA levels, this can be readily determined by examining circulating mtDNA, in particular by examining leukocyte mtDNA from samples already collected from the patient. If the level of mtDNA is measured together with the protein level of another biomarker, it is advantageous that either of these are measured on the basis of a unique sample collected from the test individual, which, if necessary, is differentially processed depending on whether a DNA or protein level measurement is performed.

[0088] More precisely, with regard to the determination of mtDNA, the "level of mtDNA" is - represents a "quantitative" value of mtDNA relative to a reference mtDNA level or a "normal" mtDNA level, in particular a value intended to represent, in particular to quantify, the amount of mtDNA in the assay sample, such as the mtDNA copy number, in particular the absolute mtDNA copy number, or - represents the relative amount of mtDNA in the tested biological sample, especially when the amount determined is normalized against the amount of nuclear DNA (nDNA) or another suitable reference also present in said biological sample, or against a normalizing gene or DNA sequence related to nDNA It can be observed that.

[0089] Thus, in certain embodiments, it is possible to determine the level of mtDNA without determining the "absolute" amount of mtDNA in a sample, but by assessing this amount by reference to another parameter.

[0090] Thus, to obtain the "level of mtDNA", techniques may be used that allow the quantification of nucleic acids from biological samples. Such techniques allow the determination of the average concentration (or amount) of nucleic acids present in a sample, i.e., mtDNA within the context of the present invention. Several methods may be used to establish such a concentration (or amount), including (1) spectrophotometric analysis of nucleic acids and their further quantification, and (2) quantification using measurements of the fluorescence intensity of dyes that bind to nucleic acids and selectively fluoresce when bound, as well as (3) quantification after specific nucleic acid amplification, such as real-time PCR, which also relies on the detection of fluorescent dyes bound to the nucleic acids to be detected and quantified. Preferably, a prior isolation of the mtDNA to be quantified may be required, according to general knowledge in the art of nucleic acid analysis.

[0091] In particular, the level of mtDNA can be determined using quantitative polymerase chain reaction (q-PCR) using routinely known protocols. Primers specific for the mitochondrial 12s RNA gene can be used, but the skilled artisan can appropriately select other genes or sequences of the mitochondrial genome to carry out such techniques, following the guidelines available in the technical literature for this technique. PCR (polymerase chain reaction) is a common method for amplifying DNA. A small amount of DNA, a DNA template, at least one pair of specific oligonucleotide primers, a suitable buffer and a thermostable DNA polymerase are required to amplify nucleotides (dATP, dCTP, dGTP, dUTP). A substance marked by a fluorophore is generally added to one of the reagents of this mixture in a thermal cycler, which contains a sensor for measuring the fluorescence of the fluorophore after excitation with the required wavelength, allowing the generation rate of one or more specific products to be measured. Real-time PCR (q-PCR) is commonly applied to the detection and quantification of DNA in samples to determine the presence and / or abundance of specific DNA sequences in these samples. Measurements are made after each amplification cycle, allowing quantification of the amplification products in real time.

[0092] Real-time PCR is carried out by using a real-time PCR machine, and after each cycle, the level of fluorescence is measured by a detector. The dyes used generally fluoresce only when bound to the DNA being amplified via PCR, and an increase in fluorescence is detected corresponding to the increasing presence of the amplification product in each amplification cycle.

[0093] Real-time PCR can be used to quantify nucleic acids by relative quantification or by absolute quantification. Absolute quantification provides the exact number of target DNA molecules by comparison with DNA standards using a calibration curve. This method allows the determination of mtDNA copy numbers in patients with pre-gastric neoplasms or suspected neoplasms and comparing this number with the mtDNA copy numbers defined in healthy subjects. (References: von Wurmb-Schwark et al., 2002, Forensic Science International, 126:34-39; Fernandes et al., 2014, Cancer Epidemiol Biomarkers Re, 23:2430-38). Relative quantification allows the determination of the fold difference between the target sequence whose amount is determined and a "housekeeping sequence".

[0094] To quantify the presence of a specific target DNA sequence, which represents the copy number of mtDNA, it is practically convenient to express its relative level in relation to another DNA sequence, called a "normalization sequence" or "housekeeping sequence", which is selected for its approximately constant expression rate. Housekeeping sequences are usually found in genes involved in functions related to basic cell survival, which usually implies constitutive gene expression. This allows the presentation of a ratio that represents the presence of amplification of the sequence of interest relative to the presence of amplification of the selected normalization agent. This method allows the assessment of the relative presence of amplification of the sequence of interest by actually knowing its absolute amount in the test sample.

[0095] Commonly used normalizing sequences are sequences found in genes that code for the following proteins: tubulin, glyceraldehyde-3-phosphate dehydrogenase, albumin, cyclophilin, ribosomal RNA sequences can be used, as a non-limiting list.

[0096] Real-time PCR allows the quantification of desired products at any time during the amplification process by measuring fluorescence. Measurements are expressed using threshold cycle (CT) values ​​(CT; the PCR cycle at which the fluorescence of the sequence of interest is detected; the lower the CT value, the more abundant the target sequence). When a normalization sequence is used to quantify the presence of target sequences, a normalization procedure such as the ΔΔCT method, which is used to analyze relative gene expression, can be used.

[0097] According to more specific embodiments, the "level of mtDNA" can be determined by quantitative polymerase chain reaction (q-PCR) via reference to the sequence of a selected normalizing gene or nDNA, and the level of mtDNA can be calculated according to the formula: ΔCt [In the formula, ΔCt=Ct nDNA -Ct mtDNA By this calculation, the level of mtDNA is calculated as the delta Ct between the mean Ct of nDNA and the mean Ct of mtDNA (ΔCT = Ct nDNA -Ct mtDNA ) can be calculated as 2ΔCt. The primers used for amplification can be selected by those skilled in the art according to the general knowledge in the art of this technique, in particular as indicated in the above referenced publications.

[0098] An example of a protocol for determining levels of mtDNA is - preparing the biological sample so as to give access to nucleic acids, in particular mitochondrial nucleic acids of the cells; - contacting the prepared sample with an oligonucleotide primer that targets mtDNA; - performing an amplification cycle; - simultaneous amplification of nDNA, which is a normalizing agent; - quantitatively detecting mtDNA and, as a normalizing agent, nDNA; - determining the level of mtDNA in light of the sequence of the selected normalizing agent, nDNA, ΔCt [In the formula, ΔCt=Ct nDNA -Ct mtDNA

[0046] The method may include a step of calculating according to the formula:

[0099] Further details regarding the measurement of mtDNA, particularly from leukocytes, are available in the Experimental Section herein; the overarching principles may also be found in WO2015 / 049372, which is incorporated herein by reference.

[0100] As shown in the experimental section of this specification, under experimental conditions, an increase in mtDNA levels above 6.3 made it possible to predict patients with AG / P with a sensitivity of 66.6% and a specificity of 65% (AUC value of 0.7089).

[0101] As shown in the experimental section of this specification, TNF-alpha protein levels were relevant in terms of the "unhealthy" decision rule, where elevated TNF-alpha concentrations were important to assess with extremely high sensitivity and very high specificity (AUC value of 0.7954) that the assay sample was derived from a non-healthy patient.

[0102] As shown in the experimental section herein, MSLN protein levels were relevant in terms of two decision rules, in particular, where a large increase in MSLN concentration was significant for identifying patients with AG / P (AUC value of 0.7433).

[0103] As shown in the experimental section herein, HP protein levels were relevant in terms of identifying patients with GC (AUC value of 0.6622).

[0104] As shown in the experimental section of this specification, SELE protein levels were relevant in terms of the "unhealthy" decision rule, where an increase in SELE concentration was significant for assessing that the assay sample was derived from a non-healthy patient, with the best AUC value of 0.7565.

[0105] As shown in the experimental section herein, the levels of the plasma protein HP were relevant as a 1.7-fold increase in the concentration of the plasma protein HP could be found in GC samples.

[0106] According to a particular embodiment, the method of the invention is a method for assessing risk, in particular a method for prognosing or diagnosing a non-atrophic gastritis (NAG) condition in a test patient according to the present description.

[0107] According to a particular embodiment, the method of the invention is a method for assessing the risk, in particular a method for prognosing or diagnosing an atrophic gastritis / preneoplastic (AG / P) condition in a test patient, according to the present description. According to this embodiment, the method of the invention is a method for assessing the risk that a human patient has atrophic gastritis / preneoplastic (AG / P), in particular the method is a method for prognosing or diagnosing an atrophic gastritis / preneoplastic (AG / P) condition in a test patient.

[0108] According to a more specific embodiment, in which an assessment for atrophic gastritis / preneoplasia (AG / P) is sought, step a. of the method of the invention consists of determining the levels of at least two, preferably between two and six, biomarkers selected from among IGFALS, KRT19, CA2, MAN2A1, KIF20B, JUP, LBP, S100A12, ATAD3B, DCD, HP, LEP, IL-8, IL-17, USF1, USF2, SELE, MSLN and EGFR, in particular determining the levels of IGFALS, KRT19, HP, LEP, MSLN and EGFR.

[0109] The biomarkers selected can be two, three, four, five, or six biomarkers.

[0110] According to multiple embodiments where an assessment for atrophic gastritis / preneoplasia (AG / P) is sought and between two and six biomarkers are used, specific embodiments of these include the selection of biomarkers as described in any one of the combinations disclosed in Table 10, which lists the best combinations of two to six biomarkers for predicting AG / P, selected based on the best AUC value obtained. As soon as four biomarkers are used according to the corresponding signature described for AG / P, the AUC is 0.82 and the sensitivity is 92%.

[0111] According to more specific embodiments where an assessment for atrophic gastritis / preneoplasia (AG / P) is sought and six biomarkers are used, these specific embodiments list the best combinations of six biomarkers for predicting AG / P: all of these combinations include the selection of biomarkers described in any one of the combinations disclosed in Table 9, corresponding to an AUC of ≧0.8, with a sensitivity between 90%-96% and a specificity between 71%-79%.

[0112] According to certain embodiments, which may be combined with any other embodiment described herein, particularly the embodiments described above, and which are methods for assessing the risk of a human patient having atrophic gastritis / preneoplasia (AG / P) for which assessment for atrophic gastritis / preneoplasia (AG / P) is sought, the method of the invention provides a method for assessing the risk of a human patient having atrophic gastritis / preneoplasia (AG / P) for at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, 100%, 101%, 102%, 103%, 104%, 105%, 106%, 107%, 108%, 109%, 200%, 210%, 220%, 230%, 240%, 250%, 260%, 270%, 280%, 290%, 300%, 310%, 320%, 330%, 340%, 350%, 360%, 370%, 380%, 390%, 400%, 410%, 420%, 430%, 440%, 450%, 460%, 470%, 480%, 490%, 500%, 510%, 520%, 530%, 540%, 550%, 560%, 570%, 580%, 590%, 600%, 610%, 620%, 630%, 640%, 650%, 660%, 670%, 680%, 690%, 700%, 710%, 720%, 98%, 99% or 100% sensitivity and / or specificity of at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100%, in particular with a sensitivity and specificity of at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100%, respectively.

[0113] According to a particular embodiment, the method of the invention is a method for assessing risk, in particular a method for prognosing or diagnosing gastric cancer (GC) status in a subject patient according to the present description.

[0114] According to a more specific embodiment, in which an assessment of gastric cancer (GC) status is sought, step a. of the method of the invention consists of determining the levels of at least two, preferably between two and six, biomarkers selected from among IGFALS, KRT19, CA2, MAN2A1, KIF20B, JUP, LBP, ARG1, S100A12, ATAD3B, DCD, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, MSLN, EGFR and STAT3, in particular determining the levels of ARG1, LEP, IL-17, TNFalpha, SELE and MSLN.

[0115] The biomarkers selected can be two, three, four, five, or six biomarkers.

[0116] According to several embodiments where an assessment of gastric cancer (GC) status is sought and between two and six biomarkers are used, specific embodiments of these include the selection of biomarkers as described in any one of the combinations disclosed in Table 10, which lists the best combinations of two to six biomarkers for predicting GC (cancer lesions), selected based on the best AUC value obtained. For GC, many of the six biomarkers allow for improved sensitivity, but for only one protein (i.e., IL-17), the specificity is already excellent (95%).

[0117] According to more specific embodiments where an assessment of gastric cancer (GC) status is sought and six biomarkers are used, these specific embodiments list the best combinations of six biomarkers for predicting GC: all of these combinations include the selection of biomarkers described in any one of the combinations disclosed in Table 11 (Table 13), corresponding to an AUC of ≧0.9, with a sensitivity of between 87% and 94%, and a specificity of between 71% and 79%.

[0118] According to certain embodiments, which may be combined with any other embodiment described herein, particularly the embodiments described above, and which are directed to assessing a gastric cancer (GC) status, the method of the present invention is a method for assessing the risk of a human patient having atrophic gastritis / preneoplasia (AG / P), and is a method for assessing the risk of a human patient having atrophic gastritis / preneoplasia (AG / P), and is directed to ... %, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% sensitivity and / or specificity of at least 80%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100%, particularly with a sensitivity and specificity of at least 80%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100%, respectively.

[0119] Regardless of whether the case is one in which assessment is sought for atrophic gastritis / preneoplastic (AG / P) status or gastric cancer (GC) status, or in general, the sensitivity and specificity can be easily determined by the established practice in the art, with the further guidance accompanying the selection of biomarkers and presented herein. The means of calculating AUC values ​​are known to those skilled in the art and have been recalled herein above. Similarly, the specificity and sensitivity, which vary depending on the cut-off value used for the determination and can be selected according to said cut-off value used for the determination, can also be defined in an optimized form using the ROC curve, also recalled herein above, and is done in the experimental section of this specification. Thus, those skilled in the art can easily adjust the biomarkers used to reach the desired and appropriate level of sensitivity and specificity.

[0120] According to other embodiments where an assessment of gastric cancer (GC) status is sought, biomarkers that are at least one biomarker selected from among CA2, KIF20B, ARG1, DCD, LEP, IL-17, TNFalpha, and MSLN have been determined to be promising (Table 3) and may be included in the signature for this purpose.

[0121] According to other embodiments where assessment for atrophic gastritis / preneoplasia (AG / P) is sought, biomarkers that are at least one biomarker selected from among IGFALS, KRT19, CA2, MAN2A1, LBP, LEP, SELE, MSLN, and EGFR have been determined to be promising (Table 3) and may be included in the signature for this purpose.

[0122] According to a particular embodiment, the method of the invention is a method for assessing the risk of a human patient having non-atrophic gastritis (NAG), or atrophic gastritis / preneoplasia (AG / P), or gastric cancer (GC).

[0123] According to a particular embodiment, the method of the invention is a method for assessing risk, in particular a method for prognosing or diagnosing a non-atrophic gastritis (NAG) condition, or an atrophic gastritis / preneoplastic (AG / P) condition, in a subject patient, in accordance with the present description.

[0124] According to a particular embodiment, the method of the invention is a method for assessing risk, in particular a method for prognosing or diagnosing a non-atrophic gastritis (NAG) status or a gastric cancer (GC) status in a subject patient, in accordance with the present description.

[0125] According to a particular embodiment, the method of the invention is a method for assessing risk, in particular a method for prognosing or diagnosing a non-atrophic gastritis (NAG) status, or a gastric cancer (GC) status, in a subject patient, according to the present description.

[0126] According to embodiments where assessment of atrophic gastritis / preneoplastic (AG / P) or gastric cancer (GC) status is sought, biomarkers that are at least one biomarker selected from among CA2, LEP, and MSLN have been determined to be promising for either status (Table 3) and therefore may be included in the signature for this purpose, including in the context of differentiation between these conditions.

[0127] According to a particular embodiment, the method of the invention is, according to the present description, a method for assessing the risk, in particular a method for prognosing or diagnosing a non-atrophic gastritis (NAG) state, or an atrophic gastritis / preneoplastic (AG / P) state, or a gastric cancer (GC) state, in a test patient, i.e. allowing the discrimination between these states. The difference between these distinct clinical situations occurs according to the value of the change observed in relation to the "normal" situation, i.e. the direction (increase or decrease) and the magnitude (fold) of the change.

[0128] According to certain embodiments, and in line with the discussion of FIG. 12, the method of the invention further comprises determining the level of at least one further protein biomarker selected from among the protein levels of LEP and S100A12, and / or the protein levels of IL-17, and / or the protein levels of LEP, S100A12, and IL-17.

[0129] It has been shown that the protein levels of LEP and S100A12 are indeed suitable for prognosing or diagnosing GC status in patients. Thus, such biomarkers, according to all possible combinations disclosed herein and added to the signature, can be used as further criteria for whether an individual whose sample is assayed may or may not have GC.

[0130] Conversely, IL-17 protein levels have been shown to be suitable for detecting healthy individuals and thus excluding gastric cancer or pre-gastric cancer conditions (see Table 4). As a result, such biomarkers, according to all possible combinations disclosed herein and added to the signature, can be used as further criteria as to whether the individual whose sample is assayed may or may not be healthy.

[0131] Such measurements may help to corroborate or refute results obtained with a biomarker signature according to the present invention.

[0132] According to one embodiment, the series of steps includes: i) implementation of the method disclosed herein, in which the assayed biomarkers comprise at least the protein levels of KIF20B and SPEN, alone or further combined with at least one other biomarker described herein, in particular the assayed biomarkers comprise at least the protein levels of KRT19, ARG1, DSP1 and HAL, and / or at least one other (or several other) protein biomarkers selected from among the protein levels of mtDNA, MSLN, HP, SEL and TNF alpha; these markers are relevant in the context of the assessment of the risk of the presence of the AG / P stage (see FIG. 12) in the assayed patient; and ii) performing the methods disclosed herein, where the biomarker being assayed is IL-17, to confirm or reaffirm whether the patient being assayed is, in fact, considered healthy (see FIG. 12); and iii) the implementation of the methods disclosed herein, in which the assayed biomarkers include at least the protein levels of LEP or S100A12; these markers are relevant in the context of the assessment of the risk of the presence of gastric cancer (see FIG. 12) in the assayed patient; such a measure can be used to confirm or reconfirm whether the assayed patient is, in fact, considered to be at risk of having gastric cancer. Inclusive of In particular, steps ii) and iii) may be performed in sequence, or only one of these steps may be performed, or only one of steps i) or ii) or iii) may be performed. A method is also disclosed.

[0133] According to an embodiment, the method of the invention is a method for assessing risk, in particular a method for prognosing or diagnosing non-atrophic gastritis (NAG) or atrophic gastritis / pre-neoplastic (AG / P), gastric cancer condition, gastric cancer (GC), or a healthy condition in a subject, in particular a method for discriminating between non-atrophic gastritis (NAG), atrophic gastritis / pre-neoplastic (AG / P), gastric cancer condition, gastric cancer (GC), or a healthy condition in a subject, in accordance with the present description.

[0134] According to a particular embodiment, the method of the present invention is a method for monitoring or diagnosing the health status of a patient predisposed to developing a gastric cancer condition, or predisposed to suffering from a precursor gastric cancer condition, or predisposed to developing a gastric cancer condition, or the health status of said patient having a pathology that puts the human patient at risk for a gastric cancer condition, which is repeated at least once over the course of time to conclude about the health status of the subject patient if the comparison set in step b. discussed herein (comparison with a "control") and / or the deviation observed in step c. discussed herein indicates a development, in particular a method for monitoring or diagnosing the health status of a patient diagnosed with gastric cancer and optionally treated for gastric cancer.

[0135] According to certain embodiments, the method is repeated as necessary, i.e. as long as necessary to monitor the evolution of the health status of a patient who may or may not be under treatment for the condition.

[0136] The method of the invention is, according to the present description, a method for assessing risk, in particular for prognosing or diagnosing a gastric cancer condition or a forestomach cancer condition, i.e., a method for diagnosing, if possible (in particular with specificity and sensitivity values, which can be easily determined by the skilled person as described herein and according to general knowledge), or, whenever required, a method for prognosticating / predicting risk in conjunction with further clues. Indeed, according to the test results, further examination of the individual may be recommended in order to fully evaluate the clinical picture.

[0137] Thus, the method of the present invention allows to determine the status of an individual's biological parameters and, if possible, to statistically determine that the individual whose biological sample is examined may present a risk of suffering from a gastric cancer condition or a forestomach cancer condition. In particular, the presence of a forestomach cancer condition, especially an AG / P condition, is associated with a risk of the presence of gastric carcinogenesis in the tested individual. Thus, further clinical investigation may be indicated. According to certain embodiments, prognosis or diagnosis requires the implementation of further clinical investigation as described herein.

[0138] According to certain embodiments, concluding that further clinical investigation is required or indicated corresponds to a conclusion about the health status of the patient from whom the test biological sample was obtained.

[0139] Therefore, "concluding on a health status" and / or "proceeding with further clinical investigation" also encompasses the enrollment of said patient in a close therapeutic monitoring procedure, i.e., said patient is recommended or directly involved in a therapeutic follow-up, including regular monitoring of the patient's condition or health status over time, and, optionally, further clinical investigation regarding his / her health status. Precisely, any determination that a subject individual is predisposed to suffer from a gastric lesion also suggests the performance of further clinical investigation. Early detection of the presence of risk for gastric carcinogenesis, i.e., detection of risk of the presence of a gastric lesion at a stage such as a stage corresponding to the AG / P state (or, alternatively, the AG / P stage), is a relevant public health goal, as detailed herein.

[0140] Non-limiting examples of "further clinical investigation" include optical examination by gastroscopy, computed tomography (or CT) scanning of the abdomen, biopsy for histological examination, various blood tests, e.g., complete blood count (CBC) to check for anemia, and other exploratory methods aimed at confirming or excluding the presence of a gastric carcinogenic process.

[0141] If a gastric lesion is suspected, "deciding as to the health status" and / or "proceeding with further clinical investigation" may include increasing the number and / or frequency of scheduled optical gastroscopic examinations, which would otherwise be performed infrequently.

[0142] If a gastric lesion is suspected and after further clinical investigation, it may also be possible to conclude, if possible, to surgically remove the area suspected of having undergone gastric carcinogenesis. In some cases, the resection may be performed by endoscopy (digestive endoscopic mucosal resection (EMR) is a procedure in which early cancers and precancerous growths are removed from the lining of the gastrointestinal tract).

[0143] Figure 16 depicts a possible procedure / scheme for the use of a diagnostic test based on the detection of preneoplastic and GC lesions by the signatures SIG-AGP and SIG-GC (described in Figure 15), respectively (see the caption for Figure 16 herein). It can be seen that the diagnostic test of the present invention can be used and repeated at different levels of the procedure presented. Thus, the method of the present invention can be used in an early diagnostic protocol and / or in a follow-up protocol, as described in Figure 16.

[0144] In a particular embodiment, the method of the invention also comprises, as a separate, simultaneous or parallel step, a step of detecting H. pylori infection, in particular via detection of an antigen specific for H. pylori infection, or via amplification of DNA and subsequent detection of said DNA, or via an assay involving detection of the presence of H. pylori-specific IgA antibodies and H. pylori-specific IgG antibodies in a biological sample taken from the test patient, or via a 13C-urea breath test performed on the test patient.

[0145] In this case, the detection of H. pylori, if any, can be carried out on a fraction (aliquot) of the biological sample taken from the patient, in particular the plasma sample or the serum fraction of the blood sample, in case of involvement, by carrying out a step of detection of an antigen specific for H. pylori infection. H. pylori-infected individuals have specific IgA and IgG antibodies, which can be easily detectable. In addition, the search for the presence of CagA antigen also confirms the presence of H. pylori. Another method for detecting H. pylori is the 13C-urea breath test (Graham et al., 1987, Lancet, 1:1174-1177), a non-invasive test with high sensitivity that is widely used in human medicine. This breath test allows the indirect measurement of H. pylori-associated urease activity. The presence of H. pylori can also be detected in stool by immunoassays indicating the presence of H. pylori antigens, or by amplification of H. pylori DNA via polymerase chain reaction (PCR), particularly using primers specific for H. pylori gene sequences available to those skilled in the art in the literature, and detection of the amplified DNA.

[0146] When both mtDNA levels and the detection of H. pylori are sought in combination, the biological sample obtained from the test individual can be a blood sample, which may on the one hand be prepared to purify a cellular fraction of the blood sample, in particular mononuclear cells or white blood cells, which contain the mtDNA to be assayed, or on the other hand to recover serum allowing the detection of H. pylori infection.

[0147] In certain embodiments, the test biological sample, particularly the plasma sample, is obtained from a patient diagnosed with gastric carcinogenesis who may or may not be under treatment for this condition, and / or a patient with an ongoing, treated or untreated H. pylori infection, and / or a patient with a history of H. pylori infection that may or may not have been eradicated by previous or ongoing treatment, and / or an individual with gastric pain and / or a family history of gastric cancer.

[0148] According to certain embodiments, which are applicable to all embodiments described herein, as detailed in this description, the levels of the plasma biomarkers are determined by enzyme-linked immunosorbent assay (ELISA) test, or mass spectrometry, or quantitative polymerase chain reaction (q-PCR), or Luminex assay, and the levels of mtDNA are determined by quantitative polymerase chain reaction (q-PCR), if performed.

[0149] Another object of the present invention is a kit suitable for carrying out the methods defined herein, or a kit for carrying out the methods defined herein, when the detection of protein levels or antigens is envisaged, comprising: at least two antibodies with different antigen specificities, each antibody being specific for a protein selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN proteins, or at least two antibodies specific for PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN proteins; A combination of several antibodies with different antigen specificities against GFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN proteins, and optionally at least one antibody specific for a H. pylori antigen, such as the CagA antigen, and optionally one or more of the following reagents: - secondary antibodies, such as biotinylated antibodies, or reagents that reveal the complexes between the specific antibodies listed above and their target(s), optionally a buffer solution, - beads such as color-coded beads, and / or magnetic or non-magnetic beads, and / or carboxylated beads, optionally with an amine coupling kit for conjugating the antibodies to the beads, - optionally phycoerythrin (PE)-conjugated streptavidin to reveal biotinylated antibodies; - optionally an assay plate, - Optionally, a notice providing instructions for use and predictions for the interpretation of the results. The present invention provides a kit comprising:

[0150] Another object of the invention is a kit suitable for carrying out the methods defined herein, when the detection of nucleic acids is envisaged, or a kit for carrying out the methods defined herein, comprising: - at least one pair of specific oligonucleotide primers or nucleic acid molecules that are specific for hybridization with mtDNA, and / or - at least two pairs of specific oligonucleotide primers or nucleic acid molecules that are specific for hybridization with a DNA region encoding two or more of the following proteins: PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN, respectively, and optionally at least one pair of specific oligonucleotide primers or nucleic acid molecules that are specific for hybridization with a nucleic acid sequence of H. pylori, and optionally one or more of the following reagents: - nucleotides (e.g. dATP, dCTP, dGTP, dUTP), - DNA polymerases, in particular thermostable DNA polymerases such as Taq DNA Polymerase, at least one dye for staining nucleic acids, in particular a dye detectable in a real-time PCR device, optionally a buffer solution, - optionally, reagents necessary for hybridization of the primers with their targets, optionally a reference dye, as well as - Notices providing instructions for use and predictive value for the interpretation of results The present invention provides a kit comprising:

[0151] Another object of the present invention is a kit suitable for carrying out the methods defined herein or a kit for carrying out the methods defined herein, according to all possible combinations of protein level, antigen and nucleic acid detection, if such combinations are envisaged, comprising: at least two antibodies with different antigen specificities, each antibody being specific for a protein selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN proteins, or at least two antibodies specific for PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN proteins; A combination of several antibodies with different antigen specificities against GFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN proteins, and optionally at least one antibody specific for a H. pylori antigen, such as the CagA antigen, and optionally one or more of the following reagents: - secondary antibodies, such as biotinylated antibodies, or reagents that reveal the complexes between the specific antibodies listed above and their target(s), optionally a buffer solution, - beads such as color-coded beads, and / or magnetic or non-magnetic beads, and / or carboxylated beads, optionally with an amine coupling kit for conjugating the antibodies to the beads, - optionally phycoerythrin (PE)-conjugated streptavidin to reveal biotinylated antibodies; - optionally an assay plate, and - Optionally, a notice providing instructions for use and a prediction for the interpretation of the results. , and - at least one pair of specific oligonucleotide primers that are specific for hybridization with mtDNA, and / or - at least two pairs of specific oligonucleotide primers, each specific for hybridization with a DNA region encoding two or more of the following proteins: PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN, and optionally at least one pair of specific oligonucleotide primers or nucleic acid molecules specific for hybridization with a nucleic acid sequence of H. pylori, and optionally one or more of the following reagents: - nucleotides (e.g. dATP, dCTP, dGTP, dUTP), - DNA polymerases, in particular thermostable DNA polymerases such as Taq DNA Polymerase, at least one dye for staining nucleic acids, in particular a dye detectable in a real-time PCR device, - optionally at least one buffer, - optionally, reagents necessary for hybridization of the primers with their targets, - Optionally, a reference dye The present invention provides a kit comprising:

[0152] Another object of the invention is a kit suitable for carrying out the method of the invention as defined herein, comprising some combination of agents or all of the agents mentioned for the kits described above, i.e. all or some of the reagents for the detection of proteins by enzyme-linked immunosorbent assay (ELISA) or for carrying out a so-called TaqMan protein assay (qPCR-based) and / or specific antibodies allowing the quantification of these proteins, if relevant, including also the positive and negative controls necessary to carry out the assay, and optionally at least one marker specific for H. pylori antigens, as well as tubes and / or means allowing the separation of leukocytes and plasma from the blood sample, in cases where the measurement of nucleic acids is mandatory (for the level of mtDNA or for the determination of the presence of H. pylori DNA or H. pylori RNA), and the reagents necessary to isolate DNA from the leukocytes and to carry out both the detection and quantification of mtDNA, including the binding of primers specific for mtDNA and nDNA genes, as relevant genes, or for H. pylori genes or RNA, as relevant genes or RNA, and also the Taq DNA required for the qPCR reaction. The goal is to actually provide a kit that includes DNA polymerase, deoxynucleotide mix, buffers, and dyes, or, according to another embodiment, all the agents and / or reagents to perform a Luminex assay.

[0153] Another object of the invention is a method suitable for carrying out the method as defined herein in any embodiment or for carrying out the method as defined herein in any embodiment, comprising at least two antibodies specific for a protein chosen from among the PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, SELE, EGFR, STAT3 and MSLN proteins, and optionally at least one pair of specific oligonucleotide primers or nucleic acid molecules specific for hybridization with mtDNA. or a set of markers comprising or consisting of at least two pairs of specific oligonucleotide primers specific for hybridization with DNA regions encoding two or more of the following proteins: PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN, and optionally at least one pair of specific oligonucleotide primers specific for hybridization with mtDNA.

[0154] The present invention also relates to the use of a kit or a set of markers as defined herein according to the present invention for screening a biological sample of blood or plasma previously taken from a human patient predisposed to suffer from a condition predisposed to developing a gastric cancer condition or predisposed to suffer from a forestomach cancer condition or predisposed to suffer from a gastric cancer condition, in particular by measuring the levels of at least two markers as defined in any embodiment herein in a biological blood or plasma sample taken from a human patient predisposed to suffer from a condition predisposed to developing a gastric cancer condition or predisposed to suffer from a forestomach cancer condition or predisposed to suffer from a gastric cancer condition, in order to determine whether a human patient has a pathology that places said patient at risk for a gastric cancer condition and / or requires further medical examinations associated therewith, or to assess the risk that a human patient will develop or that he or she has a gastric cancer condition, in particular for prognosing or diagnosing a forestomach cancer condition or a gastric cancer condition.

[0155] In a particular embodiment, the invention relates to the use of a kit or set of markers as defined herein for the screening of a blood or plasma biological sample previously taken from a human patient susceptible to suffering from a gastric cancer condition or a condition susceptible to developing into a forestomach cancer condition or a condition predisposed to developing into a gastric cancer condition, in particular for the prognosis or diagnosis of a gastric cancer condition or a forestomach cancer condition by measuring the levels of at least two markers as defined in any embodiment herein in a biological blood or plasma sample taken from a human patient susceptible to suffering from a gastric cancer condition or a condition predisposed to developing into a forestomach cancer condition or a condition predisposed to developing into a gastric cancer condition.

[0156] In accordance with the present disclosure, the use of the kit or set of markers defined herein in accordance with the present invention is to search for the parameters detailed herein and / or to monitor said parameters in a test individual as intermediate biological parameters prior to any further investigation.

[0157] The present invention also relates to the use of the agents, components or reagents disclosed herein, in particular as described in any embodiment for the manufacture of a kit suitable for carrying out the method of the present invention, or a kit for this purpose, as described herein, in particular as described in relation to a kit suitable for carrying out the method of the present invention. It may be advantageous if instructions or guidelines for use are provided for carrying out the method of the present invention and / or for obtaining a suitable kit.

[0158] According to another aspect, it is to be understood that the method of the present invention described in any embodiment herein can be implemented, at least in part, by a computer. In particular, steps b. and c. of the method of the present invention described in any embodiment herein can be implemented by a computer, where data corresponding to the levels of at least two biomarkers described in any embodiment herein are provided as output. According to another particular embodiment, the computer also measures, via suitable means of interfacing the computer and the level measuring device according to any embodiment described herein, the data collected in step a., i.e., a collection of levels of at least two biomarkers selected from among the levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, EGFR, STAT3, and mtDNA, with the proviso that the selected biomarkers do not consist of the collection of IL-8 and mtDNA levels. It may also drive in vitro harvesting.

[0159] In line with the possibility that a computer may implement steps b. and c. of the method of the invention described herein in any embodiment, the invention also provides a computer-implemented method for detecting whether a human patient has a pathology that places said patient at risk for a gastric cancer condition and / or requires further medical examination related thereto, comprising: a. receive levels of at least two biomarkers selected from among levels of PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, EGFR, STAT3, and mtDNA, with the proviso that the selected biomarkers do not consist of the set of IL-8 and mtDNA levels, or receive any level of at least two biomarkers as defined in any embodiment described herein, the determination of which is in particular in accordance with any embodiment described herein, for measuring such levels; measuring via an in vitro method or a device adapted thereto; and b. processing the level determined in step a. by comparing it to a control; and c. determining, in particular when the levels of at least two biomarkers compared in step b. deviate from their controls, via predefined decision rules, in particular decision rules related to the sensitivity and / or specificity determined in relation to the biomarkers sought, whether the levels received in step a. indicate that the human patient, in which these levels have been determined, in particular via the in vitro method for measuring the levels defined in step a., has a pathology that places said patient at risk for a gastric cancer condition and / or requires further medical examinations, in particular clinical investigations, associated therewith; The present invention also relates to a method comprising the steps of:

[0160] The determination of step c. may be performed according to any rule disclosed in the present disclosure and according to the biomarkers selected for implementation. For example, Table 2 (Table 4) herein provides an exemplary list of involvement decision rules. The decision rules may also be derived from cutoff values ​​determined for sensitivity and specificity values ​​determined to be acceptable for the test performed. The present description may rely on all examples of biomarker lists associated with sensitivity, specificity, and AUC values. Those skilled in the art may easily implement a computer-implemented method incorporating the discussed step c. based on any data described in the present disclosure. It is to be understood that the computer-implemented method discussed herein may be performed for any embodiment described herein, particularly in the context of the in vitro method for determining whether a human patient has a lesion that places the patient at risk for a gastric cancer condition and / or requires further medical testing related thereto, as defined in the claims or any embodiment described herein.

[0161] The present invention also relates to a data processing apparatus comprising a processor comprising means for performing the computer-implemented method discussed above, in particular steps b. and / or c. of said method or adapted (or configured) for carrying out said method, in particular a processor adapted (or configured) for carrying out steps b. and / or c. of said method.

[0162] According to a particular embodiment, such a data processing device comprises: - an input interface for receiving the levels of at least two biomarkers (or any combination of biomarkers described in this description) defined in step a. of the computer-implemented method discussed above, - a memory for storing at least instructions of a computer program, the memory including instructions which, when executed by a computer or processor, cause the computer to carry out the computer-implemented method discussed above, and optionally a memory for storing control data and decision rules; - a processor that reads said instructions and accesses the memory to perform the computer-implemented method discussed above; - an output interface resulting in a decision as to whether the levels of at least two biomarkers (or any combination of biomarkers described in the present description) defined in at least step a. of the computer-implemented method discussed above, render the human patient in whom these biomarker levels were measured to have a pathology placing said patient at risk for a gastric cancer condition and / or requiring further medical testing, in particular clinical investigation, associated therewith.

[0163] The invention also relates to a computer program (or computer product) comprising instructions which, when executed by a computer or a processor, cause the computer to perform the computer-implemented method discussed above.

[0164] The present invention also relates to a computer readable medium, in particular a non-transitory computer readable recording medium, having stored thereon the computer program (or computer product) discussed above, which, when the computer program (or computer product) is executed by a computer or processor, implements the computer implemented method discussed above.

[0165] The present invention is in particular a platform for non-invasive tests, which are carried out on biological samples previously obtained from an individual, which may be a patient. By non-invasive tests is meant that the method of the invention is in particular an in vitro method. Said method does not require the presence of a medical professional for its implementation. According to the present invention, several biomarkers are proposed for the early detection of gastric carcinogenesis, in particular the presence of gastric lesions involved in the gastric carcinogenesis process or at the basis of this process, for example in the AG / P stage or AG / P state as described herein. The present invention is particularly suitable for preventive purposes, but also for 1) monitoring the progression of the gastric carcinogenesis process in individuals under or not under ongoing treatment for the affected condition, and / or 2) monitoring the shift of a pre-neoplastic state to a neoplastic state, and / or 3) as a follow-up to screen for recurrence of the disease after cure. For this purpose, the method of the present invention simply relies on the detection and / or monitoring of physiological parameters of the patient.

[0166] The methods of the invention may be used for patients undergoing treatment, such as chemotherapy and / or radiation treatment, as indicators of treatment efficacy, disease stage, and disease onset.

[0167] As used herein, the term "comprising," which is synonymous with "including" or "containing," is an open-ended term and does not exclude additional, unrecited elements, components, or method steps, whereas "consisting of" is a closed term that excludes any additional elements, steps, or components not expressly recited.

[0168] The term "consisting essentially of" is an open term that does not exclude additional, unrecited elements, steps, or ingredients, so long as those additional elements, steps, or ingredients do not substantially affect the basic and novel characteristics of the application.

[0169] Thus, the term "comprising" includes the term "consisting of" as well as the term "consisting essentially of." Thus, in this application, the term "comprising" is more specifically intended as a term that encompasses the terms "consisting of" and "consisting essentially of."

[0170] In an attempt to assist the reader of this application, this description is divided into various paragraphs or sections. These separations should be considered as separating the content of one paragraph or section from the content of another paragraph or section. On the contrary, this description encompasses all combinations of the various sections, paragraphs, and sentences that may be envisioned.

[0171] The pertinent disclosures of each of the references cited herein are specifically incorporated by reference.

[0172] The features described herein above, as well as other features of the present invention, will become apparent upon reading the examples and figures, which illustrate the experiments carried out by the inventors, complementary to the features and provisions presented in this description. The following examples are presented for illustrative purposes. However, the examples are not limiting to the described invention. [Brief description of the drawings]

[0173] [Figure 1] Plasma levels of candidate biomarkers measured for all samples of the cohort: A) mtDNA measured by qPCR on DNA isolated from circulating leukocytes; C) IL-8, E) TNF-α, G) IL-17, I) USF1, and K) USF2 measured using commercial ELISA assays as described in the Methods section. Distribution of plasma levels of candidate biomarkers in the different groups of patients, H, NAG, AG / P, GC, according to the determined cut-off values: B) mtDNA, D) IL-8, F) TNF-α, H) IL-17, J) USF1, L) USF2. [Diagram 2] Plasma levels of candidate biomarkers measured for all samples of the cohort: A) LEP, C) HP, E) SELE, G) MSLN measured using commercial ELISA assays as described in the Methods section. Distribution of plasma levels of candidate biomarkers in the different groups of patients, H, NAG, AG / P, GC according to the determined cut-off values: B) LEP, D) HP, F) SELE, H) MSLN. [Diagram 3] FIG. 1 shows the diagnostic accuracy of candidate biomarkers as determined by ROC curve analysis (True Positive Rate (TPR) in function of False Positive Rate (FPR)) and AUC values. The ROC curves were obtained using the decision rule "If biomarker amount is above x (or below x or equal to or greater than x or equal to or less than x, depending on the configuration), then the patient is H or NAG or AG / P or GC," where x is the cutoff value. The optimal cutoff and AUC criteria are presented at the top left of each plot. The optimal cutoff was determined using a compromise between TPR and FPR. The table corresponding to FIG. 3 is shown below: [Table 1] [Figure 4] FIG. 4 shows the diagnostic accuracy of candidate biomarkers as determined by ROC curve analysis (True Positive Rate (TPR) in function of False Positive Rate (FPR)) and AUC values. The ROC curves were obtained using the decision rule "If biomarker amount is below x (or above x or equal to or greater than x or equal to or less than x, depending on the configuration), then the patient is H or NAG or AG / P or GC," where x is the cutoff value. The optimal cutoff and AUC criteria are presented at the top left of each plot. The optimal cutoff was determined using a compromise between TPR and FPR. The table corresponding to FIG. 4 is shown below: [Table 2] [Diagram 5]Functional association networks between potential biomarkers identified by Proteome Profiler arrays (STRING analysis: https: / / string-db.org / ) and their associated cellular functions. Known interactions: database-curated interactions (grey lines); experimentally determined interactions (--------). Predicted interactions: co-expression (·········), other interactions: text mining (black solid lines). [Figure 6] Correlation matrix analysis, hierarchical clustering analysis, and PLS-DA and sparse PLS-DA analyses. A) The paired correlation matrix represents the Pearson correlation coefficient between each pair of samples, calculated using all complete intensity value pairs measured in these samples. B) Hierarchical cluster analysis, performed as indicated in the Methods section. C) Partial least squares-discriminant analysis (PLS-DA) was used to explore the proteomic differences between the four patient groups (H, NAG, AG / P, GC). The PLS-DA plots presented a good separation between healthy patients and other pathological conditions. D) The sparse PLS-DA method selected a set of 85 potential biomarkers, allowing a clear distinction between H subjects and GC patients, but was unable to separate NAG and AG / P patients. [Figure 7] Figure 1 shows a heat map displaying the deviation from the mean intensity level of the most relevant biomarker candidates identified by MS. For most factors, a common profile is observed, mainly for NAG and AG / P, while a distinct profile characterizes GC compared to other groups. This figure contains information in color: the colored drawings are those submitted to the Patent Office at the time of filing and may be relied upon. [Figure 8]Analysis and prediction of candidate biomarkers measured by ELISA assay. A) Redistribution of AUC values ​​for all combinations of two variables for each group of patients: healthy (H), nonatrophic gastritis (NAG), preneoplastic (AG / P) and cancer (GC). B) Number of times a candidate appears in the model yielding the best AUC criterion for two variables. [Figure 9]Figure 1 shows the residual deviance of the model, measuring its ability to predict all pathologies simultaneously. The 10 best combinations of two biomarkers, highlighting the ability of the combination of IL-17 and LEP to predict H and patients (top). The 10 best combinations of three biomarkers show the confluence of TNF-α, IL-17, and LEP separating H from patients, mainly GC (bottom). The corresponding 10 best combinations of two biomarkers and 10 best combinations of three biomarkers shown in Figure 9 are reproduced below: 10 best combinations of two biomarkers (in terms of residual deviance) (top panel) 1) IL-17 and LEP: 151.23 2) TNF-α and IL-17: 154.25 3) IL-17 and USF2: 162.58 4) Gender and IL-17: 176.26 5) IL-17 and MSLN: 190.51 6) mtDNA and IL-17: 191.22 7) Hp status and IL-17: 191.22 8) IL-17 and USF1: 196.64 9) IL-8 and IL-17: 198.49 10) IL-17 and HP: 198.81The 10 best combinations (in terms of residual deviance) of the three biomarkers (see figure below) 1) TNF-α, IL-17, and LEP: 124.88 2) IL-17, USF2, and LEP: 135.12 3) mtDNA, IL-17, and LEP: 135.28 4) IL-17, HP, and LEP: 139.20 5) Gender, TNF-α, and IL-17: 140.53 6) TNF-α, IL-17, and USF2: 141.95 7) Gender, IL-17, and USF2: 142.34 8) IL-17, MSLN, and LEP: 142.98 9) TNF-α, IL-17, and MSLN: 143.40 10) mtDNA, TNF-α, and IL-17: 143.54 It is noted that the references in this figure and above to "gender" and "Hp status" refer to the fact that whether the gender or Hp status of the patient was examined in the course of the experiments performed, i.e., at each disease stage, at each detection time according to the definitions presented herein, patients known to be positive for H. pylori infection may have developed. [Figure 10]Figure 1 shows the results of model estimation: A) Distribution of AUC criteria for two and three variables for non-atrophic gastritis (NAG), pre-neoplastic (AG / P), and cancer (GC) groups. B) Number of times a candidate biomarker appears in the model yielding the best AUC criterion. [Figure 11] Figure 1 shows the residual deviance of the model, measuring its ability to predict all pathologies simultaneously. The 10 best combinations of two biomarkers, highlighting the ability of KIF20B in combination with ARG1 or CPA4 to predict GC (top). The 10 best combinations of three biomarkers show perfect classification for all groups by cluster: KRT19, KIF20B, and SPEN (bottom). The corresponding 10 best combinations of two biomarkers and 10 best combinations of three biomarkers shown in Figure 11 are reproduced below: 10 best combinations of two biomarkers (in terms of residual deviance) (top panel) 1) ARG1 and KIF20B: 21.86 2) KIF20B and CPA4: 21.95 3) DSP and KIF20B: 25.06 4) SPRR1A and S100A12: 25.17 5) S100A12 and CPA4: 25.36 6) MAN1A1 and SPRR1A: 25.55 7) ARG1 and S100A12: 25.61 8) S100A12 and MAN2A1: 25.87 9) KRT19 and S100A12: 25.95 10) CFP and CDSN: 26.04 10 best combinations of 3 biomarkers (in terms of residual deviance) (Figure below) 1) KRT19, KIF20B, and SPEN: 0.47 2) ARG1, KIF20B, and SPEN: 2.37 3) DSP, KIF20B, and SPEN: 2.38 4) HAL, KIF20B, and SPEN: 3.50 5) C7, KIF20B, and SPEN: 5.09 6) JUP, KIF20B, and SPEN: 5.79 7) KIF20B, SPEN, and CPA4: 6.01 8) SPRR1A, KIF20B, and SPEN: 6.24 9) MAN2A1, KIF20B, and SPEN: 6.68 10) KRT14, KIF20B, and SPEN: 8.49 [Figure 12]Figure 1 shows the biomarker signature that provides the most complete prediction for identifying the different stages of the GC process. As shown, IL-17 allows to distinguish between healthy subjects and patients. Among patients, the combination of KIF20B, SPEN with KRT19, ARG1, DSP1, or Hal provides a good prediction of preneoplasia (AG / P). Candidates that should be considered among other candidates according to the corresponding AUC values ​​observed for the prediction of preneoplasia are also indicated in italics. In addition, LEP and S100A12 are suitable for a good prediction of GC. [Figure 13] FIG. 1 shows plasma levels of candidate biomarkers confirmed by ELISA. Violin plots depicting plasma levels of candidate biomarkers initially identified by MS or Proteome Profiler and confirmed by commercial ELISA or measured directly by ELISA. Quantification was performed on all samples of the cohort. Statistical analysis using Mann-Whitney test with significance for p<0.05. This figure includes ELISA results for the biomarkers EGFR and STAT3. [Figure 14] Figure 1 shows a graphical representation of functional networks among potential biomarkers by STRING. Among the 22 proteins identified, 14 and 2 are functionally linked, and 2 proteins are also functionally linked. In addition, some proteins are part of physical complexes, such as STAT3-EGFR-LEP-IL-8; MSLN-LBP; and USF1-USF2 (https: / / string-db.org). [Figure 15] Figure 1 shows the best biomarker signatures for predicting A. gastric pre-neoplasia (SIG-AGP) and B. gastric cancer (SIG-GC). The AUC increases with the length of the signature for both SIG-AGP and SIG-GC. The AUC correlates with increasing sensitivity for SIG-GC and specificity for SIG-AGP. [Figure 16]Figure 1 shows an overview of the use of diagnostic tests based on the detection of preneoplastic and GC lesions by the SIG-AGP and SIG-GC signatures, respectively. Three different levels of use can be proposed: 1) screening of patients at risk of GC; 2) following the presence of preneoplasia to detect the occurrence of GC at an early stage; 3) following GC patients after surgery and during / after chemotherapy to prevent cancer recurrence.The legend of the schematic is as follows: 1. Testing a biological sample, e.g. a blood or plasma sample, taken from a patient or a pool of patients, for example for the SIG-AGP or SIG-GC signatures described herein or any other combination of markers described herein; 2. If the test of 1. is negative, no further action is envisaged; 3. If the test of 1. is positive for AG / P risk (SIG-AGP positive), proceed further to 5.; 4. If the test of 1. is positive for GC risk (SIG-GC positive), proceed further to 5.; 5. Carry out further clinical investigation, e.g. an endoscopy procedure, on the patient from whom the assay sample was taken; 6. If step 5. concludes, via the further clinical investigation of 5., the presence or risk of pre-neoplasia, proceed further to 8. as a patient follow-up procedure; 7. If step 5. concludes, via the further clinical investigation of 5., the presence or risk of gastric cancer, proceed further to 9. as a patient follow-up procedure; 8. A new / further testing step of a biological sample, e.g. a blood or plasma sample, taken from the follow-up patient, for example for the SIG-AGP or SIG-GC signature described herein or any other combination of markers described herein, which may be the start of a new round of testing starting from step 1. of this schematic diagram; 9. Further medical measures, such as treatment and / or surgery of the follow-up patient according to the data under the control of the practitioner. Step 10. A new / further testing step of a biological sample, e.g. a blood or plasma sample, taken from the follow-up patient, for example for the SIG-AGP or SIG-GC signature described herein or any other combination of markers described herein, which constitutes a patient follow-up procedure and may be the start of a new round of testing starting from step 1. of this schematic diagram. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS EXAMPLES

[0174] method Study population The study cohort is described in Table 1. It included 48 healthy (H) asymptomatic volunteers recruited at the Institut Pasteur's Clinical Investigation and Biomedical Research Support Unit (ICAReB). Each H sample was confirmed for their H. pylori negative serology using a commercially available enzyme-linked immunosorbent assay (ELISA) (Serion ELISA Classic). 26 non-atrophic gastritis (NAG) patients, 38 atrophic gastritis / preneoplasia (AG / P) patients, and 68 gastric cancer (GC) patients were included in the cohort. NAG and AG / P patients were diagnosed in the Department of Hepatology and Gastroenterology (AP-HP, A. Pare hospital, Boulogne-Billancourt), headed by Professor D. Lamarque. GC patients were diagnosed at the Department of Hepatology and Gastroenterology and Digestive Oncology, Headed by Prof. J. Taieb, AP-HP, HEGP, Paris. All patients were adults, not under anticancer treatment and not previously treated with antibiotics, bismuth compounds, proton pump inhibitors, or nonsteroidal anti-inflammatory drugs for at least 2 weeks. Diagnosis was based on endoscopy and histopathological analysis of gastric biopsies. All patients were informed and asked to sign a consent form. The study was approved by the Institut Pasteur Translational Research Center (protocol reference number: 2013-29).

[0175] [Table 3]

[0176] Clinical sample collection and histological analysis For each patient, 10 ml of blood was collected and gastric tissue specimens were isolated. Gastric biopsies were collected from both the antrum and gastric body during gastroscopy. Biopsies were immersed in formalin and processed for histological analysis and hematoxylin-eosin (H&E) staining for diagnosis of gastric lesions. The presence of H. pylori was confirmed by Giemsa staining and serology. Prior to inclusion in the cohort, the H. pylori-negative status of blood samples from healthy volunteers was verified by serological testing by the ICAReB platform.

[0177] Circulating mitochondrial DNA (mtDNA) and quantification Peripheral blood (10 ml) was collected from each patient and leukocytes were separated on Leucosep® tubes by Pancoll gradient. DNA was prepared from the separated leukocytes using the Qiamp DNA kit (Qiagen) and frozen at −80°C until tested for mtDNA quantification. In parallel, plasma fractions were also separated and frozen at −20°C until use. mtDNA levels were measured on DNA isolated from circulating leukocytes by quantitative polymerase chain reaction (q-PCR) using the StepOne™ Plus real-time PCR system and FastStart Universal SYBR Green Master (Applied Biosystems) as previously described (6) using the 12S ribosomal RNA gene and the nuclear-encoded 18S ribosomal RNA gene as endogenous references. Relative mtDNA levels were calculated as the delta Ct (ΔCt) between the mean Ct of nDNA and the mean Ct of mtDNA (ΔCt = Ct nDNA -Ct mtDNA ) to 2 ΔCt was used as the calculation.

[0178] Quantification of plasma levels of different potential biomarkers The plasma levels of the different selected candidate biomarkers: interleukin 8 (IL-8) (reference number: DY208); interleukin 17 (IL-17) (reference number: DY317); tumor necrosis factor alpha (TNF-α) (reference number: DY210); mesothelin (MSLN) (reference number: DY3265); E-selectin (SELE) (reference number: DY724); haptoglobin (HP) (reference number: DY8465); leptin (LEP) (reference number: DY398); and upstream stimulatory factor 1 and upstream stimulatory factor 2 (USF1; reference number: MBS9342772 and USF2; reference number: MBS9321077; MyBioSource) were assessed using a commercially available ELISA assay from the Duo Set RD system.

[0179] Plasma levels of factors associated with oncology pathways were screened by proteomic profiler analysis using Human XL Oncology array (reference number: ARY029; R&D systems). The array consists of the capture of selected antibodies present on a nitrocellulose membrane incubated with plasma samples mixed with a cocktail of biotinylated detection antibodies and revealed by streptavidin-horseradish peroxidase according to the supplier's recommendations. For each pathway, duplicate samples were analyzed, representing each patient group: NAG group, AG / P group, and GC group, including healthy (H) subjects.

[0180] Large-scale screening of potential plasma biomarkers using mass spectrometry (MS)-based proteomics Plasma samples In this pilot study, four groups were investigated, including H patients, NAG patients, AG / P patients, and GC patients, with n=10 samples each. Representative plasma samples from each group were selected within the same cohort from AP-HP Hospital and used for all projects.

[0181] Immunodepletion with MARS Hu-14 Plasma samples were depleted using MARS Hu-14 (5188-6560: Agilent) according to the manufacturer's protocol. Briefly, 300 μg total protein was diluted with Buffer A and filtered at 0.22 μm. Each sample was loaded onto a spin column and centrifuged at 100 g / 1 min / room temperature. After a 5 min incubation period, non-depleted proteins were eluted with Buffer A in two rounds of 400 μL each via centrifugation at 100 g / 2.5 min / room temperature. The three filtrates were combined and further precipitated with 40% (vol:vol) TCA overnight. Samples were washed twice with acetone and air-dried before being subjected to in-solution digestion.

[0182] Digestion in solution The depleted samples were diluted with 100 μL of 8 M urea / 100 mM NH 4 HCO 3 The proteins were resuspended in denaturing buffer and subjected to reduction with 5 mM TCEP (646547: Sigma, St Louis, Missouri, USA) for 15 min, followed by alkylation with 20 mM iodoacetamide (I114: Sigma, St Louis, Missouri, USA) for 30 min in the dark. Proteins were digested with 0.5 μg rLys-C (V1671: Promega, Madison, Wisconsin, USA) for 3 h at 37°C, then diluted 9-fold for subsequent digestion with 0.5 μg Sequencing Grade Modified Trypsin (V5111-Promega, Madison, Wisconsin, USA) overnight at 37°C. Digestion was stopped with 4% formic acid (FA) and peptides were desalted by reversed-phase C18 Stage-Tips method (8). Peptides were eluted with 80% acetonitrile (ACN) / 0.1% FA. Finally, samples were dried in a vacuum centrifuge and resuspended with 2% ACN / 0.1% FA. All samples were spiked with iRT peptides as recommended by Biognosys.

[0183] Peptide fractionation for spectral libraries A "pool" sample consisting of 40 plasma samples was dedicated to obtain a spectral library for the data-independent acquisition (DIA) method. The "pool" sample was depleted, digested, and peptide fractionated using poly(styrene-divinylbenzene) reversed-phase sulfonate (SDB-RPS) Stage-Tips method as described previously (8)(9). Briefly, three SDB-RPS Empore disks were stacked on a P200 chip and seven serial eluents were applied as follows: eluent 1 (60 mM ammonium formate (AmF) / 20% ACN / 0.5% FA), eluent 2 (80 mM AmF / 30% ACN / 0.5% FA), eluent 3 (95 mM AmF / 40% ACN / 0.5% FA), eluent 4 (110 mM AmF / 50% ACN / 0.5% FA), eluent 5 (130 mM AmF / 60% ACN / 0.5% FA), eluent 6 (150 mM AmF / 70% ACN / 0.5% FA), and eluent 7 (80% ACN / 5% ammonium hydroxide). All fractions were dried and resuspended with 2% ACN / 0.1% FA prior to injection. All fractions were spiked with iRT peptide as recommended by Biognosys.

[0184] mass spectrometry Data-dependent acquisition (DDA) for spectral libraries: A nanochromatography system (Proxeon EASY-nLC 1200: Thermo Fisher Scientific, Waltham, Massachusetts, USA) was online coupled to a Q Exactive™ HF mass spectrometer (Thermo Fisher Scientific) using an integrated column oven (PRSO-V1: Sonation GmbH, Biberach, Germany). For each sample, 100% solvent A (H 2After an equilibration step in 1000 Hz (0.1% FA), 1 μg of peptides was injected onto a 44 cm in-house manufactured C18 column (ReproSil-Pur Basic C18, 1.9 μm particles, 100 Å pore size: Dr. Maisch GmbH, Ammerbuch-Entringen, Germany). Peptides were eluted with a multi-step gradient over 132 min at a flow rate of 250 nL / min: 2-7% solvent B (80% ACN, 0.1% FA) over 5 min, 7-23% solvent B over 70 min, 23-45% solvent B over 30 min, and 45-95% solvent B over 5 min. The column temperature was set at 60 °C. Mass spectra were acquired using Xcalibur software using a data-dependent top 10 method with a survey scan (300-1700 m / z) at a resolution of 60,000 and an MS / MS scan (fixed first mass: 100 m / z) at a resolution of 15,000. The AGC target values ​​and maximum injection times for the survey and MS / MS scans were 3.0 × 10 6 , 100 ms, and 1.0 x 10 5 The isolation region was set to 1.6 m / z and the normalized collision energy was fixed at 28 for HCD fragmentation. 4 For an intensity threshold of 2.0×10 3 A minimum AGC target of 100% was used. Unassigned precursor ion charge states as well as charge states of 1, 7, 8, and >8 were discarded to invalidate peptide matches. Isotope exclusion was allowed and selected ions were subjected to dynamic exclusion for 45 seconds.

[0185] Data-Independent Acquisition (DIA) for Plasma Samples: Mass spectra were collected in data-independent acquisition mode using the same nanochromatographic system coupled online to a Q Exactive™ HF mass spectrometer with XCalibur software. For each sample, 100% Solvent A (H 2After an equilibration step in 0.1% FA, 1 μg of peptides was injected onto a 50 cm in-house manufactured C18 column (ReproSil-Pur Basic C18 with 1.9 μm particles and 100 Å pore size: Dr. Maisch GmbH, Ammerbuch-Entringen, Germany). Peptides were eluted with the same multi-step gradient as the spectral library. Each cycle was constructed as follows: the resolution of one MS full scan was 60000 (scan range: between 349 and 1214 m / z) and the AGC was 3.0 × 10 6 and the maximum injection time was set to 60 ms. Every MS1 was followed by an isolation region of 25 m / z, 36, covering the MS1 range. The AGC target with automatic maximum injection time was 2.0×10 5 and the NCE was set to 28. All acquisitions were performed in positive and profile modes.

[0186] Data processing for protein identification and quantification Spectral library construction: Raw data were analyzed using MaxQuant software version 1.5.0.30 (10) with the search engine Andromeda (11). MS / MS spectra were searched against the Human SwissProt database (20,203 entries as of December 4, 2018). Variable modification (oxidation of methionine and N-terminal acetylation) and fixed modification (carbamidomethylation of cysteine) were set for the search, and trypsin with a maximum of two missed cleavage was selected for the search. The minimum peptide length was set to 7 amino acids, and the false discovery rate (FDR) for peptide and protein identification was set to 0.01. The tolerance for the primary search peptide was set to 4.5 ppm, and for MS / MS match tolerance it was set to 20 ppm. Secondary peptides were allowed to identify cofragmentation events.

[0187] Data analysis for DIA: DIA experiments were analyzed using Spectronaut X (v.13.2.190705.43655; Biognosys AG). Mass tolerance at MS1 and MS2 levels was used as dynamic. XIC RT Extraction Window was set to dynamic with a correction factor of 1. Calibration method was set to automatic with nonlinear iRT calibration and precision iRT was enabled. Decoys were generated using the mutated method and dynamic limit. P-value estimation was performed using kernel density estimator. Interference correction was enabled without proteotypicity filter. Major grouping was performed by protein group id and minor grouping was performed by stripped sequence. Major group quantity was the mean peptide quantity. Major group top N was enabled with a minimum of 1 and a maximum of 3. Minor group quantity was the mean precursor quantity. Minor group top N was enabled with a minimum of 1 and a maximum of 3. Quantity MSLevel was MS2 and quantity type was area under the curve. Q value was used for data filtering. Cross run normalization was enabled with row selection as Q value sparse and normalization strategy as local normalization. Default labeling type was unlabeled with no profiling strategy and no unify peptide peaks strategy enabled. Protein inference workflow was set to automatic.

[0188] Mass spectrometry proteomic data will be deposited to the ProteomeXchange Consortium via the PRIDE partner repository ( 12 ).

[0189] statistical analysis For each group of patients, the plasma levels for each candidate biomarker were first analyzed statistically using the Mann-Whitney test. Results were considered significant when P<0.05.

[0190] Determination of false positive rate (FPR), true positive rate (TPR), ROC curve, and area under the curve (AUC) For a potential biomarker, or combination of biomarkers, a decision rule can be deduced to predict the stage of the disease (healthy, gastritis, preneoplastic, or cancer). Thus, the false positive rate (FPR) and true positive rate (TPR) can be calculated from this decision rule as follows: FPR=N / W TPR=M / C [Where N = number of patients for whom the prediction of the diagnosis was inaccurate (e.g., number of non-cancer patients for whom cancer was predicted); W = number of patients for whom the prediction of the diagnosis is inconsistent with the actual diagnosis (e.g., number of non-cancer patients); M = number of patients for whom the prediction of the diagnosis was accurate (e.g., number of cancer patients for whom cancer was predicted); C = number of patients for whom the prediction of the diagnosis is consistent with the actual diagnosis (e.g., number of patients with cancer)] It is estimated by:

[0191] When using a single potential biomarker, the stage of the disease can be predicted using a threshold value for the concentration of the potential biomarker. Thus, FPR and TPR are calculated for each threshold value (Table 2). By varying the threshold value, a receiver operating characteristic (ROC) curve is determined with FPR on the x-axis and TPR on the y-axis (Figures 3 and 4). The area under the ROC curve is a measure of the AUC. If AUC=0.5, the decision rule has no predictive power since the biomarker cannot outperform a random selection of patients (whatever the value of the threshold chosen). If AUC>0.5, the decision rule outperforms a random selection of patients (at least for one threshold chosen). The closer the AUC is to 1, the better the decision rule is (for an ideal decision rule, AUC=1). Similarly, diagnostic models are estimated for combinations of potential biomarkers to predict each stage of disease, estimating FPR and TPR and also yielding AUC values ​​(Table 4).

[0192] Exemplary references on ROC curves: Delacour, H. et al., "LaCourbe ROC (receiver operating characteristic): principles et principales applications en biologie clinique," Annales de biologie clinique, 2005, 63(2):145-54.

[0193] Multivariate data analysis and selection of potential biomarkers Correlation matrix and hierarchical clustering: Pairwise correlation analysis and hierarchical clustering were performed to highlight similarities between plasma samples. The correlation matrix represents the Pearson correlation coefficient between each pair of samples, calculated using all complete pairs of intensity values ​​measured in these samples. Hierarchical cluster analysis was performed via multiscale bootstrap resampling (1000 bootstrap iterations) with Ward's method and a correlation-based distance measure with the pvclust function in the pvclust package in R, after log2 transformation of intensities, imputation of missing values ​​with the impute.slsa function in the imp4p package in R, and normalization using the sample median centering method within conditions.

[0194] Partial least squares-discriminant analysis (PLS-DA) and sparse PLS-DA: PLS-DA was used to explore proteomic differences among the four patient groups (H, NAG, AG / P, GC). PLS-DA and sparse PLS-DA were performed using the mixOmics R package.

[0195] Predictive and diagnostic tests using combinations of two or three potential biomarkers To identify combinations of two or three potential biomarkers that allow predicting patient groups from the measured intensity (mass spectrometry) or amount (ELISA data), the following multinomial logistic regression model for combinations of k biomarkers:

[0196]

number

[0197] [In the formula, · k is the number of biomarkers used in the model. ·v i is the relative intensity value measured for biomarker i. icorresponds to the concentration or intensity measured in a gastritis sample, a preneoplastic sample, or a cancer sample divided by the average of the concentration or intensity values ​​measured in healthy patients for this protein] was used.

[0198] The model estimates were assessed using the residual deviance criterion, a goodness-of-fit statistic that assesses how well the model fits the observed stage of disease (Table 4). The model parameters, a i , b i , and c i Once estimated, the following formula:

[0199]

number

[0200] It is possible to calculate the probability that a patient has each stage of the disease.

[0201] The highest probability among these four probabilities determines the most likely state of the patient. This predicted state was compared to the actual state of the patient to determine the AUC criteria for each stage of the disease (Table 4). Thus, the estimated model can be used to diagnose the disease state of the patient.

[0202] result Candidate biomarkers selected according to their known role in the carcinogenic process mtDNA Both mtDNA mutations and variations in mtDNA content have been reported in different types of tumors. In a previous study performed on a cohort of Mexican patients, we reported high mtDNA levels (mtDNA>20) in circulating leukocyte mtDNA of GC patients, with a cutoff value of 8.23 ​​(OR: 3.93), differentiating GC samples from H samples (6). These data were similar to mtDNA quantification performed on a Moroccan patient cohort (ACIP 10-2015: Collaborative study by F. Maachi, Institut Pasteur Morocco). In the present study, the mean mtDNA levels were 1.25 (P=0.04); 1.8 (P=0.0006), and 1.3 (P=0.003) times higher in NAG, AG / P, and GC cancer patients, respectively, compared to H subjects (Figure 1A). Analysis of the distribution of samples according to mtDNA value shows that mtDNA>5 is observed in 89% of AG / P samples compared to 32% in the healthy group. More precisely, mtDNA>6.3 leads to the prediction of patients with preneoplasia (AG / P) with a sensitivity of 66.6% and a specificity of 65% (Table 2). The receiver operating characteristic (ROC) curve for the mtDNA data reports an area under the curve (AUC) value of 0.7089 for AG / P samples (Figure 3).

[0203] - Inflammatory factors IL-8, IL-17, and TNFα GC is an inflammation-driven disease. Although variations in the levels of inflammatory mediators such as IL-8, IL-17, and TNFα are not specific for the presence of gastric pre / neoplasia (AG / P), they may help to identify patients in whom a malignant process in the stomach is initiated or progressing. In our previous study of a Mexican cohort, high plasma levels of IL-8 combined with the measurement of mtDNA allowed for improved detection of GC patients (6). As indicated in the Methods section, plasma levels of IL-8, IL-17, and TNFα were assessed by commercial ELISA assays for all samples of this AP-HP cohort. As reported in Figure 1C, the mean IL-8 values ​​increased with the stage of gastric lesions, being 2 (P=0.0098), 1.6, and 3 (P<0.0001) times higher in NAG, AG / P, and GC samples, respectively, compared to H samples. It is important to note that 94% of H samples showed IL-8 ≦40 ng / ml, compared to only 39% in GC patients. In addition, IL-8 >40 ng / ml was observed in 33% and 61% of AG / P and GC samples (FIG. 1D). Calculation of cut-off values ​​according to the "unhealthy" decision rule indicates that IL-8 >17.7 pg / ml corresponds to unhealthy samples with a sensitivity of 73.2% and a specificity of 72.3% (Table 2). In addition, IL-8 >29.4 pg / ml makes it possible to predict patients with cancer (GC) with a sensitivity and specificity of 74.6% and 72.9%, respectively (Table 2).

[0204] Interestingly, IL-17 unambiguously differentiated H subjects from patients with NAG, AG / P, or GC lesions. 100 percent of NAG, AG / P, and GC samples showed IL-17 plasma levels >20 pg / ml. In contrast, IL-17 ≦20 pg / ml was observed in 100% of H samples (FIGS. 1G and 1H). More precisely, calculation of cutoff values ​​according to the "unhealthy" decision rule indicates that IL-17 >41 pg / ml corresponds to unhealthy samples with 100% sensitivity and specificity (Table 2). The favorable biomarker properties of IL-17 are also indicated by ROC curve analysis using the decision rule "if the biomarker amount is above x (or below x or ≥ x or ≤ x depending on the configuration), the patient is H or NAG or AG / P or GC" [where x is the cutoff value], with AUC values ​​of 0.75, 0.72, and 0.67 for NAG, AG / P, and GC samples, respectively (Figure 3). For healthy (H) samples, an AUC of 1 is observed using the decision rule "if the biomarker amount is below x (or above x or ≥ x or ≤ x depending on the configuration), the patient is H or NAG or AG / P or GC" (Figure 4).

[0205] A gene polymorphism of TNF-α has already been associated with an increased risk of GC (13). Measurement of plasma levels of TNF-α showed 1.6-, 1.4-, and 1.6-fold higher values ​​in NAG, AG / P, and GC samples compared to healthy subjects. Similar to IL-17, TNF-α >80 pg / ml was observed in 100%, 87%, and 93% of NAG, AG / P, and GC samples compared to 26% of H subjects (Figures 1E and 1F). More precisely, calculation of cut-off values ​​according to the "unhealthy" decision rule indicates that TNF-α >74 pg / ml corresponds to unhealthy (H) samples with both sensitivity and specificity of 98.8% and 72.3%, respectively (Table 2). Thus, using the decision rule "if biomarker amount is below x (or above x or equal to or greater than x or equal to or less than x, depending on the configuration), then the patient is H or NAG or AG / P or GC," where x is the cutoff value, an AUC value of 0.7954 is obtained (Figure 4).

[0206] Upstream stimulatory factors USF1 and USF2 USF1 and USF2 are pleiotropic transcription factors involved in the regulation of several genes associated with important cellular functions such as immune response, cell proliferation, and maintaining genomic stability (14). These factors have already been proposed as tumor suppressors (15). A recent study by our research team reported that depletion of USF1 in gastric biopsies from GC patients was associated with poor prognosis (16). Therefore, USF1 could be a potential biomarker candidate to identify patients at risk for GC.

[0207] In the present study, plasma levels of USF1 were 3.6-fold higher in GC patients compared to H subjects (P=0.0002), with 36% and 0% of GC and H samples, respectively, with USF1 >300 pg / ml (Figures 1I and 1J). In this case, the AUC value determined by ROC curve analysis, following the decision rule "if the biomarker amount is above a cutoff value x (or below x or ≥x or ≤x, depending on the configuration), the patient is H or NAG or AG / P or GC", is 0.774 for GC (Figure 3). Data obtained by measuring USF2 did not show significant differences in plasma levels between samples. However, 89%, 76%, and 77% of NAG, AG / P, and GC samples, respectively, showed USF2 >10 pg / ml, compared to only 51% for healthy individuals (Figures 1K and 1L).

[0208] Haptoglobin In plasma, haptoglobin (HP) binds to free hemoglobin. High serum levels of HP are associated with tumor progression and poor prognosis, as reported in non-small cell lung cancer (17). Serum HP has also been proposed as a novel molecular biomarker predicting liver metastasis in colorectal cancer (CRC) (18). Recently, aberrant glycosylation of serum HP has been associated with GC (19). In the present study, the mean plasma levels of HP were elevated in GC samples compared with H (1.7-fold; P = 0.0006) (Figure 2C), and 44% of GC patients showed HP levels >1.5 g / l compared with 7% for H subjects (Figure 2D).

[0209] Biomarker candidates identified by Proteome Profiler analysis Biomarker candidates were also explored by Proteome Profiler analysis, which is composed of a membrane-based antibody array that allows for the parallel determination of the relative levels of selected Human XL Oncology pathway proteins (84 cancer-related proteins), as described in the Methods section. Three candidates were selected from the Oncology pathway array, including leptin (LEP), E-selectin (SELE), and mesothelin (MSLN), according to significant variations in their plasma levels in AG / P and GC samples compared to H samples. These were then quantified in all samples from the cohort using commercially available specific ELISA assays, as indicated in the Methods section.

[0210] Leptin LEP is a particular candidate of interest due to its role as a digestive peptide hormone. LEP is an inducer of inflammatory cytokines. Its dysregulation has been reported in a wide variety of malignancies, including gastrointestinal malignancies. In CRC, its expression increases stepwise from normal mucosa to adenocarcinoma with high-grade dysplasia (20). High serum levels of leptin are associated with gastrointestinal metaplasia and increased risk of GC (21, 22). In the present study, the mean plasma levels of LEP are three-fold higher in AG / P samples compared with H and GC samples (P<0.0001). In addition, 85% of AG / P samples show LEP levels >6 ng / ml compared with only 7% in the H group (Figures 2A and 2B). LEP is a good candidate for identifying samples from patients with preneoplasia (Figure 3), as also supported by the ROC curve that determined an AUC value of 0.705 for AG / P. The cutoff calculation indicates that LEP >7.1 ng / ml corresponds to patients with preneoplasia (AG / P) with a sensitivity of 75% and a specificity of 86.2% (Table 2). In addition, LEP <4.1 ng / ml also results in the prediction of GC patients with a sensitivity of 82.3% and a specificity of 72.3%. Furthermore, low LEP values ​​of <2 ng / ml correspond to GC samples, but not H samples, with a sensitivity of 100% and a specificity of 94.9% (Table 2).

[0211] E-selectin / CD62E Selectins are glycoproteins. E-selectin (SELE) is expressed on endothelial cells under NFκB-mediated transcriptional regulation. Its expression is crucial for controlling the accumulation of leukocytes during inflammation. Plasma levels of SELE are elevated immediately upon entering the NAG stage (P<0.0001 vs. healthy) (Figure 2E), with 85%, 78%, and 84% of NAG, AG / P, and GC samples having SELE levels >8 ng / ml, compared with 24% in healthy samples (Figure 2F). Calculation of cutoff values ​​indicates that SELE >7.1 ng / ml corresponds to non-healthy samples with a sensitivity of 89.3% and a specificity of 76.3% (Table 2). As indicated in FIG. 3, using the decision rule "if biomarker amount is above x (or below x or equal to or greater than x or equal to or less than x, depending on the configuration), the patient is H or NAG or AG / P or GC" [where x is the cutoff value], the best AUC value for GC samples is 0.7565 by ROC curve analysis.

[0212] Mesothelin Mesothelin (MSLN) may be involved in cell adhesion. MSLN has been reported to be overexpressed in some human tumors (23). In our study, the mean plasma level of MSLN was 2-fold higher in AG / P samples (P < 0.0001 vs. H) (Fig. 2G), with 90% of the samples having MSLN > 10 ng / ml compared to 28% in the H group. ROC curve analysis also indicates that MSLN may be considered a useful biomarker candidate to identify patients with AG / P with an AUC of 0.7433 (Fig. 3). When MSLN is > 8.3 ng / ml, there are no samples in the healthy group, and MSLN > 10.8 ng / ml provides a prediction of AG / P patients with a sensitivity of 84.2% and a specificity of 63.1% (Table 2).

[0213] [Table 4]

[0214] Analysis of the associated functions of all these selected biomarker candidates, as reported in Figure 5, shows a link between SELE, HP, IL-8 (CXCL-8), IL-17, and TNFα, which are jointly associated with inflammatory responses. As with LEP, these have already been suggested as biomarkers in the diagnostic and prognostic assessment of pancreatitis (24). LEP, which is also linked to HP, SELE, and TNFα, is also associated with signal transduction, as are USF1 and USF2. Interestingly, MSLN has already been mentioned as a protein biomarker for ovarian cancer, as well as LEP, HP, IL-8 (CXCL-8), IL-17, and TNFα (25).

[0215] Potential biomarkers identified through large-scale screening of plasma proteins by mass spectrometry (MS)-based proteomics To obtain a large panel of candidate biomarkers and define a signature profile that could improve the earliest possible detection of patients at risk for GC, a pilot high-throughput proteomics study was developed with the UTechS MSBio platform at the Institut Pasteur. The study included the same four groups: H, NAG, AG / P, and GC, with 10 representative samples per group. Plasma sample preparation and MS analysis were performed as described in the Methods section.

[0216] After plasma depletion, data analysis of the spectral library and DIA collection, comparison between the different groups, led to the identification of a total of 224 differentially abundant proteins. When samples from NAG, AG / P, and GC groups were compared with H samples, 114, 88, and 136 proteins, respectively, showed significant variations, some of which were generally found to be differential in the comparison. According to correlation matrix analysis and hierarchical clustering analysis (see Methods section), the plasma proteome quantifications for cancer (or healthy) patients generally correlated well between themselves, but not as well with patients affected by other pathologies (Figures 6A and 6B). The plasma proteome quantifications for preneoplastic (AG / P) and NAG patients generally seem to correlate well together, but not as well with GC patients and H subjects. Partial least squares-discriminant analysis (PLS-DA) and sparse PLS-DA were used to explore proteomic differences between the four groups of samples (H, NAG, AG / P, GC). The PLS-DA plot showed good separation between H subjects and other disease states (Figure 6C). However, the PLS-DA plot could not clearly separate NAG, AG / P, and GC, indicating that the plasma proteomes of these three disease states are close and distinct from H individuals.

[0217] Sparse PLS-DA was used to select a subset of proteins that contribute to the grouping. This method selected a set of 49 potential biomarkers without missing values, which allows to clearly distinguish between healthy individuals and cancer patients, but does not allow to separate patients with preneoplasia (AG / P) and patients with gastritis (NAG) (Figure 6D), as also indicated by the heatmap reported in Figure 7, which presents the deviations from the average intensity levels of some of these candidates. Among the most involved candidates, five: IGFBP3, IGFALS, KIF20B, DCD, MAN2A1, listed in Table 3 (Table 5), make it possible to predict AG / P patients, and six: ATAD3B, DCD, S100A12, TFRC, IGHG1, CSTA are candidates for GC patients.

[0218] [Table 5-1]

[0219] [Table 5-2]

[0220] [Table 5-3]

[0221] Predictive and diagnostic tests using combinations of two or three potential biomarkers Candidate substances identified by ELISA and Proteome Profiler The same analysis was performed for candidates identified by ELISA and Proteome Profiler array. The estimated ROC curves and AUCs, specific for each pathology, for the combination of two biomarker candidates are reported in FIG. 8A. As above, for each pathology, it is possible to assess the number of times the protein appears in the model that results in the best AUC criterion (FIG. 8B), and the residual deviance of the model (FIGS. 9A and 9B). For these candidates, a perfect prediction of gastritis, preneoplasia, or cancer seems difficult, since the combination of two or three markers does not result in AUC=1. However, the combination of LEP and mtDNA results in a prediction of preneoplasia (AUC=0.761), and in the case of LEP combined with IL-17 (AUC=0.8705), a prediction of cancer. Moreover, as reported in Table 4, LEP in combination with HP, USF2, SELE, IL-8, or USF1 allows prediction of preneoplasia (0.61≦AUC≦0.70), LEP in combination with mtDNA, SELE, IL-8, or USF1 can predict GC (0.74≦AUC≦0.78), and IL-17 in combination with mtDNA (AUC=0.76) or TNFα (AUC=0.75) can also predict GC. When combinations of three candidate biomarkers are considered, LEP is always found in the five combinations predicting preneoplasia, and in four of the five it is associated with mtDNA (0.69≦AUC≦0.73). The AUC values ​​for predicting GC for the eight best combinations of the three biomarkers are good and are included between 0.84 and 0.87. For all of them, IL-17 and LEP are present, which are combined with mtDNA, HP, SELE, MSLN, IL-8, USF1, or USF2 (Table 5).

[0222] [Table 6-1]

[0223] [Table 6-2]

[0224] [Table 7]

[0225] Candidate substances identified by mass spectrometry Multinomial logistic regression models were estimated using combinations of two or three biomarkers among the candidates identified by MS analysis, listed in Table 3. Two approaches were used to evaluate the results of the model estimation. The first approach was to use the estimated model to predict the pathology from the combination of biomarkers and to estimate the ROC curves and AUC values ​​specific to each pathology (Figure 10A). Then, for each pathology, it is possible to assess the number of times the protein appears in the model, resulting in the best AUC criterion (Figure 10B). The second approach was to calculate the residual deviance of the model, which corresponds to the estimated modeling error, and allows to measure the ability of the model to predict all pathologies simultaneously (Figures 11A and 11B).

[0226] Our results suggest that it is difficult to completely predict preneoplasia and gastritis using only two biomarkers (no combination of biomarkers achieves AUC=1 for these conditions), whereas patients suffering from cancer can be completely predicted using 30 combinations of two biomarkers (Figure 10B), including KRT14, CFP, ARG1, SA10012, ATAD3B, KIF20B, SPEN, SERPINA5, DSP, CPA4, KRT19, JUP, KRT2, CDSN, MAN1A1, MAN2A1, SPRR1A, HAL, and DCD, as reported in Table 6. It is important to note that S100A12 is the most frequently selected candidate among the 30 combinations detected by two variables to predict GC. In addition, patients with gastritis (NAG) could be predicted (0.8≦AUC≦0.85) using 10 combinations of two biomarkers, including PGK1, CFP, IGFALS, KRT19, CPA4, CA2, SERPINA5, and MAN2A1, among which 10 combinations of two biomarkers, including PGK1, CFP, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1, LBP, and DSP (Table 6), could predict patients with preneoplasia (AG / P). Using three biomarkers, all pathologies could be predicted perfectly. Very interestingly, the combination of three biomarkers that resulted in the lowest residual deviation: KRT19, KIF20B, and SPEN; ARG1, KIF20B, and SPEN; DSP, KIF20B, and SPEN resulted in perfect prediction (AUC=1) of all pathologies for NAG, AG / P, and GC. KRT19, KIF20B, and SPEN have special suitability as cancer biomarkers. Keratin 19 (KRT19) plays an important role in different types of cancer and is used as a prognostic marker (26)(27). Kinesin family member 20B (KIF20B) may promote cancer development due to its effects on cell proliferation and apoptosis. As an example in hepatocellular carcinoma, its high expression is associated with advanced tumor stage and poor prognosis (28).In addition, Msx2-interacting protein (SPEN) has been suggested as a novel tumor suppressor and regulates the Notch pathway (29). As shown in Table 3, KIF20B and SPEN, as well as JUP, SERPINA5, and F13A1, are among the cancer-associated plasma proteins.

[0227] [Table 8-1]

[0228] [Table 8-2]

[0229] [Table 9-1]

[0230] [Table 9-2]

[0231] Confirmation of MS data by ELISA Further studies were performed to confirm the ability of the biomarker candidates identified by MS analysis to predict preneoplasia (AG / P) or GC. Among the proteins listed herein, the plasma levels of at least 11 proteins were quantified by ELISA for all samples of the cohort, as indicated in Table 3, where the protein names are also mentioned.

[0232] As reported in Figure 13, significant differences are observed for plasma levels of ARG1, JUP, MAN2A1, LBP, and IGFALS, including the previously measured LEP, for both AG / P and GC groups compared to H group. Plasma levels of DCD are significantly lower in AG / P compared to H and GC. In addition, levels of ATAD3B, HP, and CA2 showed significant differences between GC and H samples (Figure 13). ROC analysis yielded AUCs of >0.6 for predicting preneoplasia (AG / P) for LEP (AUC=0.685), ARG1 (AUC=0.63), and HP (AUC=0.629), and AUCs of >0.7 for predicting GC for ARG1 (AUC=0.712) (Table 3).

[0233] Among these candidates, ARG1 is a key element of the urea cycle that catalyzes the conversion of arginine to ornithine and urea, which are further metabolized to proline and polyamides, and drives collagen synthesis and bioenergetic pathways. ARG1 is also involved in the modulation of immune responses to cancer, and elevated levels have been reported in the tumor microenvironment of GC (30). In our study, ARG1 was able to predict AG / P with a sensitivity (Sens) of 72% and a specificity (Spec) of 54%, while in the case of GC, the sensitivity was 49% but the specificity was 93%.

[0234] According to STRING analysis (Figure 14), a functional link is indicated by JUP, also known as γ-catenin. JUP is involved in cell-cell junctions as focal adhesions and in tight junctions. JUP is involved in cytoskeletal rearrangements. Its loss correlates closely with GC malignancy and poor prognosis (31). In this example, we observed an overall increase in plasma levels of JUP in both NAG, AG / P, and GC samples, but the AUC for predicting GC was only 0.5, with low sensitivity (27%) and specificity (73%).

[0235] Functionally linked to JUP (Figure 14), HP binds to free hemoglobin. Recently, abnormal glycosylation of serum HP has been associated with GC (9). In the present study, HP >1.5 g / l was observed in 44% of GC patients compared with 7% of H subjects. However, the AUC for predicting GC was only 0.567, with a specificity of 69% and a sensitivity of 44%.

[0236] LBP is functionally linked to LEP (Figure 14). LBP is a glycoprotein that binds to various bacterial LPS and plays a role in innate immune responses. We observed a significant increase in plasma levels of LBP from the NAG to GC stages (Figure 13). LBP shows AUCs of 0.5 and 0.562 predicting AG / P and GC, respectively. Note that in 100% of H samples, LBP is ≦250 ng / ml, whereas in 73% and 79% of AG / P and GC samples, respectively, LBP is >250 ng / ml.

[0237] DCD can be cleaved into several peptides with different functions. Its best known role is as an anti-mitochondrial host defense protein. Dysregulation of DCD expression has been reported in a variety of cancers, including GC (32). The AUC values ​​determined for DCD to predict AG / P and GC were 0.597 and 0.581, respectively, with sensitivity: 88%; specificity: 31% for AG / P and sensitivity: 22%; specificity: 94% for GC.

[0238] ATAD3B plays a role in mitochondrial network organization in stem cells and has been shown to be re-expressed in cancer cells (33). Plasma levels of ATAD3B were higher in GC samples compared to healthy samples (Figure 13), with an AUC for GC of 0.583.

[0239] CA2 contributes to pH regulation in the upper duodenal villous epithelium during proton-coupled peptide absorption. Overexpression of CA2 has been reported in gastrointestinal mesenchymal tumors (34). Higher levels of CA2 were observed in plasma from GC patients compared to H subjects (Figure 13), but the AUC was only 0.583.

[0240] Two other potential biomarkers, which were also quantified, are MAN2A1 and IGFALS. MAN2A1 is involved in the biosynthesis of glycans, of which different types have been reported in the serum of GC patients (35). MAN2A1 is associated with an AUC of ≦0.5 for both AG / P and GC. The last one, IGFALS, is a serum protein that binds insulin growth factor (IGF). IGFALS has already been suggested as a marker for malignant progression in liver cancer (36). Similar to MAN2A1, IGFALS is associated with an AUC of 0.5 predicting both AG / P and GC.

[0241] In addition to these candidates, we also included in our panel factors that are amenable to strong biomarkers due to their known roles in inflammation and cancer, as mentioned above for IL-8, TNFα, USF1, and USF2. STAT3 and EGFR, also listed in Table 3, are two of these additional candidates that were measured by ELISA on all samples of the cohort.

[0242] STAT3 is a key transcription factor involved in cellular responses to interleukins, LEP, and other growth factors (Figure 14). It is important to note that STAT3 has been previously reported to be upregulated in GC (37). Phosphorylation of STAT3 participates in the LEP signaling pathway and contributes to LEP resistance, a major risk factor for obesity (38). In this example, we observed that plasma levels of STAT3 were elevated in GC patients (Figure 13). All and 89% of H and NAG samples, respectively, showed STAT3 levels of ≦250 ng / ml, whereas in the majority (59%) of GC samples, STAT3 was >5 ng / ml, with an AUC value of 0.672 (sensitivity: 62.8%; specificity: 71.6%).

[0243] The second biomarker is EGFR, which plays a pivotal role in cell proliferation and tumor development. EGFR is overexpressed in 27-64% of gastric tumors and has been proposed as an indicator of poor outcome in GC patients (39). Elevated EGFR levels were observed in GC samples compared to AG / P samples (Figure 13), with AUCs of only 0.5 and 0.472 for AG / P and GC, respectively.

[0244] Predictive and diagnostic tests using combinations of two to six candidate biomarkers The present data as well as the literature on biomarker discovery argue for the importance of a combination of biomarker panels rather than a single candidate substance to improve early detection in cancer patients. Indeed, our results suggest that it may be difficult to fully predict preneoplasia and / or cancer using a single biomarker, since the best AUC values ​​for single candidates to predict GC are 0.731 and 0.725 for IL-17 (sensitivity: 50.5%, specificity: 95.6%) and IL-8 (sensitivity: 63%, specificity: 82%), respectively. For preneoplasia, LEP yields the best AUC value: 0.685 (sensitivity: 94.2%, specificity: 42.8%). To improve detection of the presence of preneoplastic or cancerous lesions, a multinomial logistic regression model was estimated using a combination of up to six biomarkers among the candidates whose plasma levels were measured by ELISA for all samples of the cohort.

[0245] Based on these data, two approaches were used to evaluate the results of model estimation. The first approach is to use the estimated model to predict pathology from a combination of biomarkers and estimate the ROC curve and AUC value specific to each pathology. Then, for each pathology, it is possible to assess the number of times that a protein appears in the model that results in the best AUC criterion. The second approach is to calculate the residual deviance of the model, which corresponds to the estimated modeling error, and allows to measure the ability of the model to predict all pathologies simultaneously.

[0246] Using these models, we were able to identify the two best signatures, including up to six biomarkers, for predicting the presence of gastric preneoplasia (SIG-AGP) and the presence of cancer lesions (SIG-GC) according to AUC, sensitivity and specificity values.

[0247] The so-called "SIG-AGP" signature corresponds to a combination of MSLN, HP, LEP, KRT19, IGFALS, and EGFR that predicts preneoplasia with an AUC value of 0.852, a sensitivity of 91.4%, and a specificity of 79% (Figure 15A).

[0248] The so-called "SIG-GC" signature includes IL-17, ARG1, LEP, MSLN, TNFα, and SELE, which predicts GC with an AUC value of 0.928, a sensitivity of 92.9%, and a specificity of 92.7% (Figure 15B).

[0249] LEP and MSLN are both found in. The median, minimum, and maximum plasma levels for each biomarker are reported in Table 8.

[0250] It is important to note that the predictive properties of these two signatures increase with the number of biomarkers. For both signatures, the AUC increases from one to six proteins, as well as the specificity for SIG-AG / P and the sensitivity for SIG-GC. For SIG-AGP, a sensitivity of 94% is already observed with LEP as the single biomarker, whereas the specificity remains at 42.8%. Similarly, for SIG-GC, the highest specificity (95.6%) is obtained with the single biomarker IL-17.

[0251] For completeness, Table 9 reports the list of the best combinations of six biomarkers for predicting AG / P. These combinations correspond to an AUC of ≥ 0.8, with a sensitivity between 90% and 96% and a specificity between 71% and 79%.

[0252] In addition, as reported in Table 10, a combination of five biomarkers: MSNL, LEP, KRT19, MAN2A1, and IGFALS could predict preneoplasia with a higher sensitivity (94.1%) and specificity of 73.7% than SIG-AGP.

[0253] In addition, to predict GC, a superior score was also observed for the combination: TNFα, IL-17, MSLN, LEP, KIF20B, ARG1 (AUC: 0.928; sensitivity 93%, specificity 92.7%), where SELE is replaced by KIF20B, as compared to SIG-GC. Note that the best specificity for GC (95.6%) was already obtained with a single biomarker, IL-17 alone, but with a lower AUC and sensitivity (0.731 and 50.6%).

[0254] For completeness, Table 11 reports the list of the best combinations of six biomarkers for predicting GC, corresponding to an AUC of ≥ 0.9 with a sensitivity between 87% and 94% and a specificity between 88% and 94%.

[0255] [Table 10]

[0256] [Table 11-1]

[0257] [Table 11-2]

[0258] [Table 12]

[0259] [Table 13-1]

[0260] [Table 13-2]

[0261] [Table 13-3]

[0262] [Table 13-4]

[0263] In addition to the 10 biomarkers that make up SIG-AGP or SIG-GC, 9 additional candidates, namely CA2, MAN2A1, ATAD3B, S100A12, USF2, JUP, CPA4, KRT14, and F13A1, are also promising because they are also found in combinations of 4 to 6 biomarkers that yield the 10 best AUCs for predicting preneoplasia or GC considering ELISA data for CA2, MAN2A1, ATAD3B, S100A12, USF2, JUP, or mass spectrometry data for CPA4, KRT14, and F13A1, an overview of which is presented below in Table 12.

[0264] [Table 14]

[0265] conclusion In conclusion, our analysis led us to a short list panel of 40 biomarker candidates, listed in Table 3, that allows predicting whether a patient suffers from gastritis, gastric preneoplastic lesions, or cancerous lesions. Some candidates provide better predictive results than others. It is important to note that the data from the ELISA analysis show that, at a first level, IL-17 can make it possible to distinguish healthy subjects from patients with good reliability. Among the patient group, the most striking results indicate that LEP in combination with IL-17 can predict GC with a high level of reliability, and LEP associated with HP or MSLN allows the detection of preneoplasia (AG / P). LEP has been validated by three methods developed in this study, and it is important to note that IL-17, SELE, MSLN, and HP have been validated by two of these.

[0266] In addition, according to the MS-based analysis performed, the combination of KIF20B and SPEN could provide good prediction for AG / P or GC. As summarized in Figure 12, their combination with KRT19 and S100A12 seems to perfectly predict GC. These promising analyses allow us to propose a mathematical model that predicts the presence of gastric lesions based on a combination of biomarkers, resulting in a first biomarker signature that detects the presence of gastric pre-neoplastic and cancerous lesions, including, but not limited to, IL-17, LEP, HP, MSLN, KIF20B, SPEN, KRT19, and S100A12, in a specific embodiment. KIF20B, SPEN, KRT19, and S100A12 were identified by MS proteomics studies including 10 samples per group, but further exploration by ELISA assays on all plasma samples of the cohort will help confirm their accuracy as biomarkers, as well as other potential candidates found in combination with KIF20B and SPEN listed in Table 6.

[0267] According to another embodiment, among the proteins initially identified by MS, the plasma levels measured by ELISA for 12 of them confirmed the MS data and led to their consideration as potential biomarkers for the detection of gastric lesions at different stages of the GC cascade. These 12 proteins, in addition to the 10 already selected candidates also quantified by ELISA, led us to define the best combination of 6 biomarkers, which in a particular embodiment correspond to two signatures, SIG-AGP and SIG-GC, predicting the presence of pre-neoplastic and GC lesions, respectively. It should be emphasized that the majority of the 10 candidates included in SIG-AGP and SIG-GC confirmed the data initially collected, including MSLN, HP, KRT19, TNFα, IL-17, LEP, ARG1, KIF20B. The predictive score of SIG-GC is excellent. In the case of SIG-AGP, the AUC, sensitivity, and specificity are slightly lower than those obtained by SIG-GC, but are still very good. The lowest specificity observed for SIG-AGP (79%) can be attributed to the heterogeneity of this sample group, which includes atrophic gastritis (AG), intestinal metaplasia (IM), and dysplasia (D). In this patient group, associations between all the different lesions were found: AG, AG+IM, AG+IM+D. A good prediction for these lesions is crucial due to the short period between the transformation of IM to D, which immediately precedes the development of cancer lesions. The lesions that are most difficult to identify by endoscopy are also these preneoplastic lesions.

[0268] In conclusion, according to this embodiment, two biomarker signatures, SIG-AGP and SIG-GC, have been identified and validated using three different approaches. These signatures constitute an important tool for predicting the presence of gastric pre-neoplasia and cancer lesions based on a simple blood sampling, paving the way for the future development of non-invasive diagnostic tests to improve the detection / prevention of GC patients. As illustrated in Figure 16, this test can be proposed not only for initial screening before driving patients to further clinical investigations, such as endoscopy, in case of positive results, but also useful for following up patients already detected to have pre-neoplasia, as well as recurrence / remission of patients under anti-cancer treatment.

[0269] Outlook The characterization of a biomarker signature that allows predicting the presence of NAG, AG / P and GC lesions paves the way for the future development of non-invasive prognostic / diagnostic tools for the early detection and prevention of individuals at risk of developing GC. This diagnostic test can be based on an ELISA assay, according to a particular embodiment, performed on plasma samples, according to a particular embodiment, combining the measurement of different factors that constitute a biomarker signature that allows predicting the presence of gastric pre-neoplastic and cancerous lesions. According to another embodiment, such a diagnostic test can be performed using a technique such as Luminex assay. The simultaneous measurement of the plasma levels of each factor that constitutes this signature and its comparison with the corresponding predefined cut-off values ​​indicates the presence or absence of a pre-neoplastic or cancerous lesion. This important tool, based on a simple blood sampling, allows the initial screening of patients at risk of GC, driving them to further clinical investigations. In addition, this diagnostic tool is also extremely useful for predicting the recurrence / outcome of the disease and monitoring the individualized treatment and follow-up of patients.

[0270] References TIFF2024502253000027.tif242170TIFF2024502253000028.tif248170TIFF2024502253000029.tif53170

Claims

1. 1. An in vitro method for indicating whether a human patient has a pathology that places said patient at risk for a gastric cancer condition and / or requires further medical examination in connection therewith, comprising screening a biological sample of blood or plasma previously taken from a human patient susceptible to a condition that may develop into a gastric cancer condition, or susceptible to a forestomachic cancer condition, or susceptible to a gastric cancer condition, determining the levels of at least two biomarkers selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, and mtDNA levels, with the proviso that the selected biomarkers do not consist of the set of IL-8 and mtDNA levels; determining the levels of at least two biomarkers selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, EGFR, STAT3, and mtDNA levels, with the proviso that the biomarkers determined do not consist of the set of IL-8 and mtDNA levels; and b. comparing the level determined in step a. to a control. The method includes:

2. wherein step a. comprises two biomarkers, IL-8 protein and mtDNA levels, optionally in combination with one or more biomarkers selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-17, TNF alpha, USF1, USF2, SELE, and MSLN; 2. The method of claim 1, comprising determining the levels of at least three biomarkers, including two biomarkers, further combined with one or more biomarkers selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-17, TNFalpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN.

3. 3. The method of claim 1 or 2, wherein step a. comprises determining the levels of at least three biomarkers, including two biomarkers selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, SELE, MSLN, and mtDNA levels, and at least one further biomarker selected from among EGFR and STAT3.

4. 4. The method of claim 1, wherein step a. comprises determining the levels of at least three biomarkers selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, EGFR, STAT3, and mtDNA levels.

5. 3. The method of claim 1 or 2, wherein step a. comprises determining the levels of at least two, preferably between two and six, biomarkers selected from among IGFALS, KRT19, CPA4, CA2, MAN2A1, KIF20B, JUP, F13A1, LBP, KRT14, ARG1, S100A12, ATAD3B, DCD, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, SELE, MSLN, EGFR, and STAT3.

6. 6. The method according to any one of claims 1 to 5, which is a method for assessing the risk of a human patient having non-atrophic gastritis (NAG), or atrophic gastritis / preneoplasia (AG / P), or gastric cancer (GC).

7. 7. The method according to any one of claims 1 to 6, which is a method for assessing a human patient for the risk of having atrophic gastritis / preneoplasia (AG / P).

8. 8. The method of claim 7, wherein step a. consists of determining the levels of at least two, preferably between two and six, biomarkers selected from among IGFALS, KRT19, CA2, MAN2A1, KIF20B, JUP, LBP, S100A12, ATAD3B, DCD, HP, LEP, IL-8, IL-17, USF1, USF2, SELE, MSLN and EGFR, in particular determining the levels of IGFALS, KRT19, HP, LEP, MSLN and EGFR.

9. 9. The method according to claim 7 or 8, which is a method for assessing the risk of a human patient having atrophic gastritis / preneoplasia (AG / P) with a sensitivity of at least 80% and / or a specificity of at least 80%, in particular a sensitivity and specificity of at least 80%, respectively.

10. The method according to any one of claims 1 to 6, which is a method for assessing the risk of a human patient having gastric cancer (GC).

11. 11. The method of claim 10, wherein step a. consists of determining the levels of at least two, preferably between two and six, biomarkers selected from among IGFALS, KRT19, CA2, MAN2A1, KIF20B, JUP, LBP, ARG1, S100A12, ATAD3B, DCD, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, MSLN, EGFR and STAT3, in particular determining the levels of ARG1, LEP, IL-17, TNFalpha, SELE and MSLN.

12. The method according to claim 10 or 11, which is a method for assessing the risk of a human patient having gastric cancer (GC) with a sensitivity of at least 75% and / or a specificity of at least 90%, in particular with a sensitivity and specificity of at least 90%, respectively.

13. 13. The method of any one of claims 1 to 12, which is a method for identifying the presence of a non-atrophic gastritis (NAG) state, an atrophic gastritis / pre-neoplastic (AG / P) state, a gastric cancer (GC) state, or a healthy state in a test patient.

14. 14. The method according to any one of claims 1 to 13, for monitoring the health status of a patient predisposed to suffering from a condition predisposed to developing a gastric cancer condition, or predisposed to suffering from a precursor gastric cancer condition, or predisposed to suffering from a gastric cancer condition, or the health status of said patient having a pathology that places said human patient at risk for a gastric cancer condition, said method being repeated at least once over time, in particular a method for monitoring the health status of a patient diagnosed with gastric cancer and optionally treated for gastric cancer.

15. 15. The method according to any one of claims 1 to 14, comprising, as a separate, simultaneous or parallel step, a step of detecting H. pylori infection, in particular via detection of an antigen specific for H. pylori infection, or via amplification of DNA and subsequent detection of said DNA, or via an assay involving detection of the presence of H. pylori-specific IgA antibodies and H. pylori-specific IgG antibodies in a biological sample previously taken from the test patient, or via a 13C-urea breath test performed on the test patient.

16. 16. The method according to any one of claims 1 to 15, wherein the blood sample is derived from a patient diagnosed with gastric carcinogenesis, optionally under treatment for this condition, and / or from a patient with an ongoing, treated or untreated Helicobacter pylori infection, and / or from a patient with a history of Helicobacter pylori infection, which may or may not have been eradicated, and / or from an individual with gastric pain and / or a family history of gastric cancer.

17. 17. The method of any one of claims 1 to 16, wherein the level of the plasma biomarker is determined and performed by an enzyme-linked immunosorbent assay (ELISA) test, or mass spectrometry, or quantitative polymerase chain reaction (q-PCR), or a Luminex assay, and the level of mtDNA is determined by quantitative polymerase chain reaction (q-PCR).

18. Further, step c: c. If the levels of at least two biomarkers determined and compared in steps a. and b. deviate from their control levels, it is concluded that the human patient has a pathology that places said patient at risk for a gastric cancer condition and / or requires further medical examination, in particular clinical investigation, associated therewith.

18. The method of any one of claims 1 to 17, comprising:

19. A kit suitable for carrying out the method as defined in any one of claims 1 to 18, comprising at least two antibodies with different antigen specificities, each antibody being specific for a protein selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN proteins, or at least two antibodies specific for PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN proteins; A combination of several antibodies with different antigen specificities against GFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN proteins, and optionally at least one antibody specific for a H. pylori antigen, such as the CagA antigen, and optionally one or more of the following reagents: - secondary antibodies, such as biotinylated antibodies, or reagents that reveal the complexes between the specific antibodies listed above and their target(s), - beads such as color-coded beads, and / or magnetic or non-magnetic beads, and / or carboxylated beads, optionally with an amine coupling kit for conjugating the antibodies to the beads, - optionally phycoerythrin (PE)-conjugated streptavidin to reveal biotinylated antibodies; optionally a buffer solution, - optionally an assay plate, - Optionally, a notice providing instructions for use and predictions for the interpretation of the results. Kit including:

20. A kit suitable for carrying out the method as defined in any one of claims 1 to 18, comprising - at least one pair of specific oligonucleotide primers that are specific for hybridization with mtDNA, and / or - at least two pairs of specific oligonucleotide primers, each specific for hybridization with a DNA region encoding two or more of the following proteins: PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN, and optionally at least one pair of specific oligonucleotide primers or nucleic acid molecules specific for hybridization with a nucleic acid sequence of H. pylori, and optionally one or more of the following reagents: - nucleotides (e.g. dATP, dCTP, dGTP, dUTP), - DNA polymerases, in particular thermostable DNA polymerases such as Taq DNA Polymerase, at least one dye for staining nucleic acids, in particular a dye detectable in a real-time PCR device, optionally a buffer solution, - optionally, reagents necessary for hybridization of the primers with their targets, optionally a reference dye, as well as - Notices providing instructions for use and predictive value for the interpretation of results Kit including:

21. A kit suitable for carrying out the method as defined in any one of claims 1 to 18, comprising at least two antibodies with different antigen specificities, each antibody being specific for a protein selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, EGFR, STAT3 and MSLN proteins, or PGK1, CFP, A combination of several antibodies with different antigen specificities against IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, EGFR, STAT3, MSLN proteins, and optionally at least one antibody specific for a H. pylori antigen, such as the CagA antigen, and optionally one or more of the following reagents: - secondary antibodies, such as biotinylated antibodies, or reagents that reveal the complexes between the specific antibodies listed above and their target(s), - beads such as color-coded beads, and / or magnetic or non-magnetic beads, and / or carboxylated beads, optionally with an amine coupling kit for conjugating the antibodies to the beads, - optionally phycoerythrin (PE)-conjugated streptavidin to reveal biotinylated antibodies; optionally a buffer solution, - optionally an assay plate, and - Optionally, a notice providing instructions for use and predictions for the interpretation of the results. 、 and - at least one pair of specific oligonucleotide primers that are specific for hybridization with mtDNA, and / or - at least two pairs of specific oligonucleotide primers, each specific for hybridization with a DNA region encoding two or more of the following proteins: PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN, and optionally at least one pair of specific oligonucleotide primers or nucleic acid molecules specific for hybridization with a nucleic acid sequence of H. pylori, and optionally one or more of the following reagents: - nucleotides (e.g. dATP, dCTP, dGTP, dUTP), - DNA polymerases, in particular thermostable DNA polymerases such as Taq DNA Polymerase, at least one dye for staining nucleic acids, in particular a dye detectable in a real-time PCR device, - optionally at least one buffer, - optionally, reagents necessary for hybridization of the primers with their targets, - Optionally, a reference dye Kit including:

22. A set of markers suitable for carrying out the method as defined in any one of claims 1 to 18, comprising or consisting of at least two antibodies specific for a protein selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNFalpha, USF1, USF2, SELE, EGFR, STAT3 and MSLN proteins, and optionally at least one pair of specific oligonucleotide primers or nucleic acid molecules specific for hybridization with mtDNA, or and a set of markers comprising or consisting of at least two pairs of specific oligonucleotide primers specific for hybridization with DNA regions encoding two or more of the following proteins: PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, EGFR, STAT3, and MSLN, and optionally at least one pair of specific oligonucleotide primers or nucleic acid molecules specific for hybridization with mtDNA.

23. 23. Use of a kit or a set of markers according to any one of claims 19 to 22 for screening a biological sample of blood or plasma previously taken from a human patient predisposed to suffer from a condition predisposed to developing into a gastric cancer condition or predisposed to suffer from a forestomach cancer condition or predisposed to suffer from a gastric cancer condition, in particular for indicating whether a human patient has a pathology placing said patient at risk for a gastric cancer condition and / or requires further medical examinations related thereto, by measuring the levels of at least two markers as defined in any one of claims 19 to 22 in a biological blood or plasma sample taken from a human patient predisposed to suffer from a condition predisposed to developing into a gastric cancer condition or predisposed to suffer from a forestomach cancer condition or predisposed to suffer from a gastric cancer condition.

24. 1. A computer-implemented method for determining whether a human patient has a condition that places the patient at risk for a gastric cancer condition and / or requires further medical testing associated therewith, comprising: a. receiving levels of at least two biomarkers selected from among PGK1, CFP, IGFALS, KRT19, SPRR1A, CPA4, CA2, SERPINA5, MAN2A1, KIF20B, SPEN, JUP, KRT6C, CDSN, KPRP, F13A1, SAA1 (SAA2), LBP, DSP, KRT2, KRT14, ARG1, S100A12, ATAD3B, MAN1A1, HAL, DCD, C7, HP, LEP, IL-8, IL-17, TNF alpha, USF1, USF2, SELE, MSLN, EGFR, STAT3, and mtDNA levels, with the proviso that the selected biomarkers do not consist of the set of IL-8 and mtDNA levels, or receiving any level of at least two biomarkers as defined in any one of claims 1 to 11, the determination of which is particularly preferred, and the levels of at least two biomarkers are selected from the set of in vivo assays for measuring such levels; measuring via an in vitro method or a device adapted thereto; and b. processing the level determined in step a. by comparing it to a control; and c. determining, in particular when the levels of at least two biomarkers compared in step b. deviate from their controls, via predefined decision rules, in particular decision rules related to the sensitivity and / or specificity determined in relation to the biomarkers sought, whether the levels received in step a. indicate that the human patient, in which these levels have been determined, in particular via the in vitro method for measuring the levels defined in step a., has a pathology that places said patient at risk for a gastric cancer condition and / or requires further medical examinations, in particular clinical investigations, associated therewith; The method includes:

25. A data processing apparatus comprising means for performing the method according to claim 24, in particular steps b. and / or c. of the method according to claim 24, or comprising a processor adapted / configured to perform the method according to claim 24, in particular a processor adapted / configured to perform steps b. and / or c. of the method according to claim 24.

26. - an input interface for receiving the levels of at least two biomarkers defined in step a. of claim 24; - a memory for storing at least instructions of a computer program, the memory including instructions that, when the program is executed by a computer or a processor, cause the computer to carry out the method according to claim 24, and optionally a memory for storing control data and decision rules; - a processor accessing a memory to read said instructions and to perform the method of claim 24; - an output interface resulting in a decision whether the levels of at least two biomarkers as defined in step a. as defined in claim 24 indicate that the human patient in whom the levels of these biomarkers have been measured has a pathology that places said patient at risk for a gastric cancer condition and / or requires further medical examination, in particular a clinical investigation, related thereto.

26. A data processing apparatus according to claim 25, comprising:

27. 25. A computer program comprising instructions which, when executed by a computer or processor, cause the computer to carry out the method according to claim 24.

28. A computer readable medium, in particular a non-transitory computer readable recording medium, having stored thereon a computer program as claimed in claim 27, which implements in particular the method of claim 24 when the computer program is executed by a computer or processor.

Images