Use of a marker set for determining the risk of an individual to have hepatocellular carcinoma

Abstract

The present invention relates, amongst others, to the use of a marker set comprising at least two substances in an in vitro method for determining the risk of an individual to have hepatocellular carcinoma. A first of the at least two substances is chosen from a first group consisting of glutamine, proline, valine, pyruvate, glycine, glutamate, hydroxybutyrate, alanine, and formic acid, and wherein a second of the at least two substances is chosen from the first group, from a second group, or from a third group, wherein the second group consists of lactate and glucose, and wherein the third group consists of high-density lipoprotein, apolipoprotein A1, low-density lipoprotein, apolipoprotein B-48, apolipoprotein B-100, triglycerides, total cholesterol, LDL cholesterol, and HDL cholesterol.

Description

[0001] Use of a marker set for determining the risk of an individual to have hepatocellular carcinoma

[0002] Description

[0003] The present invention relates to the in-vitro use of a marker set for determining the risk of an individual to have hepatocellular carcinoma according to the preamble of claim 1 , to the further medical use of such a marker set according to the preamble of claim 13, and to an analysis method for determining the risk of an individual to have hepatocellular carcinoma according to the preamble of claim 14.

[0004] Hepatocellular carcinoma (HCC) is an aggressive tumor that mainly develops in patients with chronic liver disease and cirrhosis. The malign tumor develops out of the liver tissue itself thereby counting as one of the primary liver cancer diseases. The growth ranges from solitary, multicenter, to diffuse infiltrative. Affected patients typically present with generalized symptoms such as jaundice, ascites, loss of appetite, fatigue, and mild to moderate upper abdominal pain often in late disease stages. Therefore, HCC is frequently diagnosed late in its course, due to the absence of pathognomonic symptoms (1 ). Seven out of ten diagnosed HCCs are already in the advanced stage when first diagnosed. Consequently, many patients have disease which is untreatable.

[0005] Worldwide, HCC represents the fifth most common cancer and the second most common cause of cancer related deaths in men (2-4). In women, it is the seventh most diagnosed cancer and the sixth leading cause of cancer death. In some countries in Southeast Asia and Africa, it is the most common malign tumor. The number of deaths per year in HCC is virtually identical to the incidence throughout the world, underscoring the high case fatality rate of this aggressive disease (5). The 5-year survival of patients with HCC is only 0 to 10 % when the tumor is diagnosed in an advanced stage (6). In contrast, 5-year survival rates of 50 to 70 % are possible when the tumor is diagnosed at an early stage, for which more curative therapies are available. Thus, the central goal of HCC surveillance programs is to detect HCC tumors as early as possible to be able to provide curative treatments and improve patient outcomes.

[0006] There is evidence that surveillance programs can reduce HCC related mortality by 37 % (7). Guidelines recommend enrolment of patients at risk, i.e., patients with liver cirrhosis or chronic hepatitis B, into regular surveillance programs. All guidelines recommend abdominal ultrasound every six months as the method of choice for HCC surveillance, but they differ in their recommendations regarding the use of the tumor biomarker alpha-fetoprotein (AFP) (8- 10). In 2009, a meta-analysis including 13 studies showed that the most widely used surveillance test, abdominal ultrasound, is able to detect HCC with a sensitivity and specificity of 94 % before tumors present with clinical symptoms (1 1 ). However, the sensitivity of ultrasound for detecting early-stage HCC was only 63 %. A more recent meta-analysis confirmed the suboptimal performance of ultrasound for early HCC detection with a sensitivity of less than 50 % (12). A major drawback of ultrasound in HCC surveillance is that its performance in detecting early HCC highly depends on the expertise of the operator and the quality of the equipment.

[0007] AFP is the most widely tested biomarker in HCC, however, as a serological test for surveillance, AFP has a suboptimal performance. One reason is that only a small proportion of tumors at an early stage (10-20 %) present with abnormal AFP serum levels (13-15). The combination of ultrasound and AFP hardly increased sensitivity for early HCC compared to ultrasound alone (69 % vs. 63 %) in one meta-analysis, while a more recent meta-analysis suggests that AFP may significantly improve sensitivity for early HCC detection (50 % vs. 63 %) for early HCC detection (11 , 12).

[0008] Other proposed serum biomarkers, such as AFP-L3 (an isoform of AFP) and DCP (des-y- carboxyprothrombin), are markers of advanced tumor stage and their performances, despite being more sensitive than AFP, are not optimal for the detection of early HCC (16-20).

[0009] Recently, a statistical model evaluating above mentioned biomarkers together with age and sex of the patient was developed to improve detection of HCC in patients with chronic liver disease, and was named GALAD based on its input variables: Gender, Age, AFP-L3, AFP, des-y-carboxy-prothrombin (21 ). The GALAD score was validated across several cohorts from Germany, UK, Japan, Hong Kong, and the U.S., showing high accuracy in detecting HCC with sensitivities greater 90 % (22-24). Despite these very promising results, the GALAD score and its modifications are not yet recommended by clinical guidelines and still warrant extensive validation of their utility in prospective studies and clinical routine. Taken together, there is still a substantial clinical need for biomarkers to complement ultrasound in the detection of HCC, in particular early HCC.

[0010] Various studies searched for specific metabolomic markers indicative for HCC in blood or urine (25-44). Several biomarkers have been shown to be influenced by the HCC tumor environment and are amenable to measurement with nuclear magnetic resonance. The hypoxic environment of HCC causes an upregulation of proline synthase and reductase, resulting in altered serum proline (45). Enhanced rates of proline synthesis is also noted in tumors to support their rapid growth and proliferation (46). The strongly upregulated glutaminolysis of rapidly proliferating HCC cells causes a significant conversion of glutamine to glutamate, strongly altering their serum concentrations (47). The increased energy demands of the rapidly dividing HCC cell results in the alteration of metabolites directly and indirectly involved in energy production via the citric acid cycle and beta-oxidation of fatty acids, namely citrate, acetylcarnitine, hydroxybutyrate, pyruvate, and mannose (48-51 ). Additionally, amino acids such as glycine and valine are consumed in the cell-building process and are also shown to oncogenic roles in HCC such as DNA-methylation and mTROC pathway activation (52) goes here. High and low-density lipoproteins (HDL and LDL) are significantly altered in HCC given their primary metabolism taking place in liver tissue and the integral role of cholesterol in cell membrane synthesis of the rapidly proliferating tumor cell (53).

[0011] WO 2022 / 240891 A1 describes metabolites that are detectable in saliva and are useful for indicating disease pathology or a breakdown in liver function. Specifically, the relative abundance of acetophenone, octadecanol, lauric acid, 3-hydroxybutyric acid, 1 -monopalmitin, 1 -monostearin and combinations thereof was found to distinguish healthy individuals from those having cirrhosis and / or HCC. Accordingly, the disclosure provides methods for evaluating the HCC status of a subject and providing treatment appropriate for the HCC status of the subject.

[0012] US 2014 / 0134751 A1 describes a method for diagnosing hepatic cancer by analyzing a sample and determining the level of at least one compound selected from the group consisting of glycine, trimethylamine-N-oxide, hippurate and citrate and comparing the levels in the sample with control levels. The methods disclosed are useful in distinguishing between a patient having hepatic cancer and a patient having cirrhosis.

[0013] US 2021 / 0208147 A1 describes methods for assessing the risk of a patient, in particular with chronic liver disease, to develop primary liver cancer over time, using functions combining blood biochemical markers, wherein the markers comprise a2-macroglobulin (A2M), GGT (gamma-glutamyl transpeptidase), haptoglobin, and apolipoprotein A-l (apoA1 ).

[0014] It is an object of the present invention to provide novel methods and biomarkers for the diagnosis of hepatocellular carcinoma as well as for determining the risk of an individual to have hepatocellular carcinoma. This object is achieved with the in-vitro use of a marker set having the claim elements of claim 1. Such a marker set comprises or consists of at least two (e.g., 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1 , 12, 13, 14, 15, 16, or 17) substances that are chosen from specific groups. The marker set necessarily comprises one substance that is chosen from a first group. The first group consists of glutamine, proline, valine, pyruvate, glycine, glutamate, hydroxybutyrate, alanine, and formic acid. The marker set further comprises at least one further substance that is also chosen from the first group or that is chosen from a second group or a third group. The second group consists of lactate and glucose. The third group consists of high-density lipoprotein (HDL), apolipoprotein A1 , low-density lipoprotein (LDL), apolipoprotein B-48, apolipoprotein B-100, triglycerides, total cholesterol, LDL cholesterol, and HDL cholesterol. LDL cholesterol refers to that part of cholesterol that is contained within low-density lipoprotein. HDL cholesterol refers to that part of cholesterol that is contained within high-density lipoprotein.

[0015] For this purpose, the concentration of the substances contained in the marker set is determined in a body fluid obtained from a patient. This concentration determination can be carried out by any appropriate measuring or analysis method, such as nuclear magnetic resonance (NMR) spectroscopy, mass spectrometry, high-performance liquid chromatography (HPLC), infrared spectroscopy such as Fourier-transform infrared (FT-IR) spectroscopy, clinical chemistry, and immunodiagnostics.

[0016] An alteration of concentration of at least two substances of the marker set with respect to the concentration in a control group or an alteration of a concentration ratio between at least two substances with respect to the concentration ratio of the same substances in a control group was correlated in a statistically significant way with the presence of hepatocellular carcinoma at the time of analysis (i.e., enabling a distinction between individuals not having hepatocellular carcinoma and individuals having hepatocellular carcinoma).

[0017] When testing individual biomarkers of the substances contained in the marker set, no significant results could be obtained for distinguishing the two groups (individuals not having hepatocellular carcinoma versus hepatocellular carcinoma patients). Rather, the area under the curve (AUC) values of receiver operating characteristic (ROC) plots showed values mainly lying below 0.75. The AUC value of ROC plots is an aggregated metric that evaluates how well a logistic regression model classifies positive and negative outcomes at all possible cut-offs. It can range from 0 to 1 .0. An AUC value of 0 represents a prediction of the opposite of the trained correlation. An AUC value of 0.5 represents a random prediction. An AUC value of higher than 0.5 represents a classification of an event as fulfilling the trained correlation, wherein higher values represent better classification. The marker sets were tested against a training dataset and a test dataset. After all training and test processes, glutamine, proline, valine, pyruvate, glycine, glutamate, hydroxybutyrate, alanine, and formic acid turned out to be valid biomarkers for the underlying question (i.e., distinguishing HCC patients from individuals not having hepatocellular carcinoma), provided that the concentration of at least two of these biomarkers was determined at the same time (i.e., in one or more body fluid samples from the same patient obtained at the same time point). Optionally, one of the before-mentioned substances can be replaced or supplemented by lactate, glucose, high-density lipoprotein, apolipoprotein A1 , low-density lipoprotein, apolipoprotein B-48, apolipoprotein B-100, triglycerides, total cholesterol, LDL cholesterol, and HDL cholesterol, since these substances also turned out to have an altered concentration in HCC patients compared to that of individuals not having hepatocellular carcinoma.

[0018] The concentration determination can be made with a method being able to determine the concentration of the substances by a single measurement or by a method requiring more than one measurement for such determination. NMR spectroscopy is particularly appropriate for such a concentration determination since it enables a highly accurate concentration determination of multiple substances in a body fluid by a single measurement in a very short measuring time.

[0019] In an embodiment, the concentration of the substances is standardized to the concentration of another substance that is naturally present in the sample. This other substance may also be listed in the first group, the second group or the third group of substances that can make up the marker set. In an embodiment, this other substance does not belong to the first, the second or the third group as defined above.

[0020] In an embodiment, glutamine and glutamate are not chosen at the same time as substances for the marker set.

[0021] In an embodiment, the marker set comprises or consists of glutamine and proline. Such a marker set showed an ALIC value of 0.83 (confer Figure 1 A).

[0022] In an embodiment, the marker set comprises or consists of glycine and proline. Such a marker set showed an AUC value of 0.83 (confer Figure 2A).

[0023] In an embodiment, the marker set comprises or consists of glutamine and valine. Such a marker set showed an AUC value of 0.83 (confer Figure 3A). In an embodiment, the marker set comprises or consists of glutamine and pyruvate. Such a marker set showed an ALIC value of 0.84 (confer Figure 4A).

[0024] In an embodiment, the marker set comprises or consists of glutamine and glycine. Such a marker set showed an AUC value of 0.83 (confer Figure 5A).

[0025] In an embodiment, the marker set comprises or consists of glutamine and formic acid. Such a marker set showed an AUC value of 0.83 (confer Figure 6A).

[0026] In an embodiment, the marker set comprises or consists of alanine and formic acid. Such a marker set showed an AUC value of 0.83 (confer Figure 7A).

[0027] In an embodiment, the marker set comprises or consists of hydroxybutyrate and formic acid. Such a marker set showed an AUC value of 0.83 (confer Figure 8A).

[0028] In an embodiment, the marker set comprises or consists of hydroxybutyrate and pyruvate. Such a marker set showed an AUC value of 0.83 (confer Figure 9A).

[0029] In an embodiment, the marker set comprises or consists of proline and glucose. Such a marker set showed an AUC value of 0.83 (confer Figure 10A).

[0030] In an embodiment, the marker set comprises or consists of glutamine and glucose. Such a marker set showed an AUC value of 0.83 (confer Figure 11 A).

[0031] In an embodiment, the marker set comprises or consists of pyruvate and lactate. Such a marker set showed an AUC value of 0.83 (confer Figure 12A).

[0032] In an embodiment, the marker set comprises or consists of pyruvate and HDL (or pyruvate and apolipoprotein A1 ). Such a marker set showed an AUC value of 0.84 (confer Figure 13A).

[0033] In an embodiment, the marker set comprises or consists of pyruvate and HDL cholesterol. Such a marker set showed an AUC value of 0.84 (confer Figure 14A).

[0034] In an embodiment, the marker set comprises or consists of pyruvate and glucose. Such a marker set showed an AUC value of 0.83 (confer Figure 15A). In an embodiment, the marker set comprises or consists of formic acid and lactate. Such a marker set showed an ALIC value of 0.83 (confer Figure 16A).

[0035] In an embodiment, the marker set comprises or consists of glutamine, proline, and pyruvate. Such a marker set showed an ALIC value of 0.84 (confer Figure 17A).

[0036] In an embodiment, the marker set comprises or consists of glutamine, proline, and glycine. Such a marker set showed an AUC value of 0.83 (confer Figure 18A).

[0037] The currently most widely used HCC biomarker AFP showed an AUC value of only 0.76 (confer Figure 19A). Thus, all of the above-mentioned two-substance of three-substance marker sets show significantly better AUC values than AFP did on the same dataset.

[0038] In an embodiment, the marker set only comprises or consists of biomarkers contained in the first group. In an embodiment, the marker set only comprises or consists of biomarkers contained in the first group or the second group. In an embodiment, the marker set only comprises or consists of biomarkers contained in the first group or the third group.

[0039] In an embodiment, the marker set comprises at least one substance chosen from the third group and being different from high-density lipoprotein and apolipoprotein A1 if the marker set comprises any of high-density lipoprotein and apolipoprotein A1 as one of the at least two substances. The concentrations of HDL and apolipoprotein A1 in a body fluid are typically closely interrelated with each other so that both substances can be exchanged against each other for many applications.

[0040] In an embodiment, the marker set comprises at least one substance chosen from the third group and being different from low-density lipoprotein, apolipoprotein B-48, apolipoprotein B- 100, and total cholesterol if the marker set comprises any of low-density lipoprotein, apolipoprotein B-48, apolipoprotein B-100, and total cholesterol as one of the at least two substances. The concentrations of low-density lipoprotein, apolipoprotein B-48, apolipoprotein B-100, and total cholesterol in a body fluid are typically closely interrelated with each other so that these substances can be exchanged against each other for many applications.

[0041] In an embodiment, the first substance is glutamine. In an embodiment, the first substance is proline. In an embodiment, the first substance is glycine. In an embodiment, the marker set additionally comprises at least one of alpha-fetoprotein, alpha-fetoprotein isoform L3, and des-y-carboxyprothrombin. The inventors found out that supplementing the marker set with one of these generally known biomarkers even increases the performance of an HCC analysis done with the marker set.

[0042] In an embodiment, the method, in which the marker set is used, comprises determining a stage of the hepatocellular carcinoma on a predeterminable scale. Thus, the marker set is, in this embodiment, not only used for determining the risk that an individual has a hepatocellular carcinoma, but also aims in classifying the severity of the hepatocellular carcinoma. Various staging scales are known that can be used for staging the severity of a hepatocellular carcinoma, such as the HCC Milan stage, the Eastern Cooperative Oncology Group (ECOG) stage, the Barcelona Clinic Liver Cancer (BCLC) stage, and - indirectly since directed to the liver function and not to the hepatocellular carcinoma itself - the Child-Pugh score and the Child-Pugh class. Any of these scales (as well as other scales) is well suited for performing staging of the detected hepatocellular carcinoma in the individual based on an analysis of a body fluid with respect to the substance concentration in the marker set.

[0043] In an embodiment, the marker set is used for detecting a hepatocellular carcinoma in an early stage such as the early Milan stage, the BCLC stages 0 and / or A, or the performance status (PS) PS 0 or PS 1 of the ECOG scale.

[0044] In an aspect, the present invention relates to the further medical use of a defined marker set for in-vivo diagnostics of hepatocellular carcinoma. The marker set comprises or consists of at least two (e.g., 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1 , 12, 13, 14, 15, 16, or 17) substances that are chosen from specific groups. The marker set necessarily comprises one substance that is chosen from a first group. The first group consists of glutamine, proline, valine, pyruvate, glycine, glutamate, hydroxybutyrate, alanine, and formic acid. The marker set further comprises at least one further substance that is also chosen from the first group or that is chosen from a second or a third group. The second group consists of lactate and glucose. The third group consists of high-density lipoprotein (HDL), apolipoprotein A1 , low-density lipoprotein (LDL), apolipoprotein B-48, apolipoprotein B-100, triglycerides, total cholesterol, LDL cholesterol, and HDL cholesterol.

[0045] In an aspect, the present invention relates to a method for analyzing an isolated body fluid sample in vitro, comprising the steps explained in the following. This method is carried out on an isolated body fluid sample originating from an individual. In a first step, the concentration of least two (e.g., 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1 , 12, 13, 14, 15, 16, or 17) substances is determined by analyzing the body fluid sample with a suited measuring technique. A first of the at least two substances is chosen from a first group consisting of glutamine, proline, valine, pyruvate, glycine, glutamate, hydroxybutyrate, alanine, and formic acid. A second of the at least two substances is chosen from the first group, from a second group, or from a third group, wherein the second group consists of lactate and glucose, and wherein the third group consists of high-density lipoprotein (HDL), apolipoprotein A1 , low- density lipoprotein (LDL), apolipoprotein B-48, apolipoprotein B-100, triglycerides, total cholesterol, LDL cholesterol, and HDL cholesterol.

[0046] Afterwards, a score is calculated from the determined concentrations, wherein the score is indicative for the risk of the individual to have hepatocellular carcinoma.

[0047] The score can be calculated by taking into consideration the concentrations measured or expected in a body fluid sample from a control group. To give a simple example, the score can be the median of the concentration ratios of the at least two substances between the body fluid test sample of the patient and corresponding control values of a body fluid control sample that have been measured in the past. If the score is above or below a predetermined threshold value, a significant increase or decrease of the marker substances is present in the body fluid test sample that is indicative for hepatocellular carcinoma. It should be noted that other calculation methods as well as a weighting of individual marker concentrations with respect to other marker concentrations can also be performed in an embodiment.

[0048] A suited way to calculate the score is disclosed on pages 25 to 27 of WO 2012 / 045773 A9. Another suited way to calculate the score is the following: wherein n a) = a + > bx• IxX=1 a = const. bx= substance specific coefficient

[0049] I = parameter being indicative for the concentration of substance x

[0050] Thereby, the individual factors a, b need to be adjusted according to the underlying model and can vary in dependence on the specific substances considered in the marker set. Parameter “I” can be, e.g., the signal intensity or signal integral of an according signal observed in the evaluated measuring result. To give an example, “I” can be the signal intensity or signal integral of an NMR signal in an NMR spectrum if NMR spectroscopy is used as measuring technique.

[0051] In an embodiment, “I” is a ratio between two signal intensities or two signal integrals. In such a case, it is, e.g., possible to standardize the concentration of a first substance (or a plurality of substances) by the concentration of a second substance.

[0052] The score is a (semi-)quantitative measure for the likelihood that the individual has developed a hepatocellular carcinoma. Thus, the score serves for (semi-)quantitatively determining the risk of presence of hepatocellular carcinoma. Alternatively or additionally, the score is a (semi--)quantitative measure for staging the hepatocellular carcinoma.

[0053] Calculating the score comprises multiplying each of the concentrations of the substances by a substance-specific weighting factor to provide a plurality of weighted values and combining the weighted values into a risk equation. Afterwards, an output of the risk equation is compared to a predefined threshold. If the score is above the threshold, there is a likelihood that the individual has developed a hepatocellular carcinoma. In an embodiment, the likelihood and / or the risk is higher, the higher the score is (i.e., the likelihood and / or the risk increases with increasing distance of the score from the threshold).

[0054] In an embodiment, the calculated score is output and presented to the individual and / or to a third person such as a physician or medical staff. The output can be performed on a display (i.e., in an electronic way) or in printed form. Thereby, it is also possible to generate a report indicating the score, optionally in combination with a comparative scale of possible scores and their meaning with respect to the risk of having a hepatocellular carcinoma or with respect to staging of a (suspected) hepatocellular carcinoma.

[0055] In an embodiment, the method is a computer-implemented method. In particular, all steps of spectral analysis and concentration determination as well as of score calculation are performed on a computer. Such steps are far too complex to be done in a manual way. The computer- implemented concentration determination is, in an embodiment, based on a spectral analysis, such as an analysis of NMR spectra. The spectral analysis and the further required steps until the score can be output can be done on the same computer that is used for controlling a spectrometer performing the spectral analysis or on a different computer. In an embodiment, the body fluid sample is a urine sample or a blood sample. In an embodiment, the blood sample is a whole blood sample, a blood serum sample, a blood plasma sample, or any other blood preparation derivable from whole blood or from other blood preparations. Blood serum is a particularly appropriate body fluid for carrying out the method or for the above-mentioned in vitro or in vivo uses of specific biomarkers contained in the marker set. Blood plasma is also an appropriate body fluid. In an embodiment, a substance of the second group of substances is present if blood plasma is used as body fluid to be analyzed. The substances contained in the second group are particularly helpful for determining highly accurate concentration values of any of the substances in the marker set, in particular of the substances of the first group.

[0056] In an embodiment, the body fluid sample (and therewith the patient from whom the body fluid sample originates) is grouped into one of at least two predefined groups based on the calculated score. Typically, one group encompasses patients suffering from hepatocellular carcinoma, wherein the other group encompasses individuals not having hepatocellular carcinoma. The resulting grouping can also be indicated on an according report. In an embodiment, the grouping encompasses more than two groups (yes / no), namely staging information on the grade of severity of the detected hepatocellular carcinoma.

[0057] In an embodiment, the individual of whom the body fluid is analyzed, is a healthy individual. In an embodiment, the individual is a risk patient, i.e., an individual belonging to a group that has an increased risk of developing a hepatocellular carcinoma with respect to the risk of a healthy standard population. In an embodiment, the individual suffers from (etiology-independent) cirrhosis. In an embodiment, the individual suffers from a (in particular chronic) viral liver infection and has a cirrhosis. In an embodiment, the individual suffers from chronic hepatitis B and has a cirrhosis. In an embodiment, the individual suffers from chronic hepatitis C and has a cirrhosis. In an embodiment, the individual suffers from non-alcoholic fatty liver disease (NAFLD), in particular from non-alcoholic steatohepatitis (NASH), and has a cirrhosis. In an embodiment, the individual suffers from alcoholic liver disease and has a cirrhosis. In an embodiment, the individual has a cryptogenic cirrhosis. In an embodiment, the individual suffers from a (in particular chronic) viral liver infection without having a cirrhosis. In an embodiment, the individual suffers from chronic hepatitis B without having a cirrhosis. In an embodiment, the individual suffers from grade IV fibrosis (i.e., a fibrosis that led to cirrhosis). All of the precedingly mentioned embodiments can be particularly well combined so that the individual may be chosen from any of the beforementioned groups of patients / healthy individuals. In an embodiment, the score is calculated by not only considering the concentration of the at least two marker substances but also includes the age and / or the gender of the individual who donated the body fluid sample.

[0058] In an embodiment, calculating the score involves calculating a ratio between at least two concentration values. To give an example, a ratio between glutamine and proline or between proline and glutamine or between glutamine and glutamate is calculated in an embodiment. Generally, individual ratios of pairs of two of all marker substances the concentrations of which have been determined upon carrying out the method can be calculated for calculating the score.

[0059] In an aspect, the present invention relates to a medical method for diagnosing hepatocellular carcinoma in an individual. The method comprises the steps explained in the following.

[0060] In a first step, a body fluid sample is gathered from an individual. In a second step, the concentration of at least two (e.g., 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1 , 12, 13, 14, 15, 16, or 17) substances is determined by analyzing the body fluid sample with a suited measuring technique. A first of the at least two substances is chosen from a first group consisting of glutamine, proline, valine, pyruvate, glycine, glutamate, hydroxybutyrate, alanine, and formic acid. A second of the at least two substances is chosen from the first group, from a second group, or from a third group, wherein the second group consists of lactate and glucose, and wherein the third group consists of high-density lipoprotein (HDL), apolipoprotein A1 , low- density lipoprotein (LDL), apolipoprotein B-48, apolipoprotein B-100, triglycerides, total cholesterol, LDL cholesterol, and HDL cholesterol.

[0061] Afterwards, a score is calculated from the determined concentrations, wherein the score is indicative for determining the risk of the individual to have hepatocellular carcinoma.

[0062] In a further aspect, the present invention relates to a decision support system for analyzing an isolated body fluid sample in vitro, the decision support system comprising: a) a unit for providing a body fluid sample from an individual; b) a unit for determining the concentration of at least two substances in the body fluid sample by analyzing the body fluid sample with a suited measuring technique, wherein a first of the at least two substances is chosen from a first group consisting of glutamine, proline, valine, pyruvate, glycine, glutamate, hydroxybutyrate, alanine, and formic acid, and wherein a second of the at least two substances is chosen from the first group, from a second group, or from a third group, wherein the second group consists of lactate and glucose, and wherein the third group consists of high-density lipoprotein (HDL), apolipoprotein A1 , low-density lipoprotein (LDL), apolipoprotein B-48, apolipoprotein B- 100, triglycerides, total cholesterol, LDL cholesterol, and HDL cholesterol; and c) a unit for calculating a score from the determined concentrations, the score being indicative for determining the risk of the individual to have hepatocellular carcinoma.

[0063] In an embodiment, the unit for determining the concentration of the at least two substances is configured to determine the concentration of any substance combination of the embodiments explained above.

[0064] While some of the explained uses and methods are described as in vitro uses and methods and some of the explained uses and methods are described as in vivo uses and methods, it should be noted that each in vitro use or method can also be carried out as in vivo use or method, and vice versa.

[0065] All embodiments of the use of the marker set can be combined in any desired way and can be transferred either individually or in any arbitrary combination to the further medical use of the marker set as well as to the different methods and to the decision support system. Likewise, all embodiments of the further medical use of the marker set can be combined in any desired way and can be transferred either individually or in any arbitrary combination to the use of the marker set, to the different methods, and to the decision support system. Finally, all embodiments of the different methods can be combined in any desired way and can be transferred either individually or in any arbitrary combination to the use of the marker set, to the further medical use of the marker set, to any other of the described methods, and to the decision support system.

[0066] Further details of aspects of the present invention will be explained in the following making reference to exemplary embodiments and accompanying Figures. In the Figures:

[0067] Figures 1 A to 18B show ROC plots (always in Figure A) illustrating the ability of different marker sets for determining the risk of an individual to have hepatocellular carcinoma as well as confusion matrices (always in Figure B) illustrating the distributions of calculated scores from the determined concentrations of the biomarkers contained in the respective marker set; and

[0068] Figures 19A and 19B show an ROC plot (in Figure 19A) illustrating the ability of alphafetoprotein for determining the risk of an individual to have hepatocellular carcinoma as well as a confusion matrix (in Figure 19B) illustrating the distribution of the calculated scores from the determined concentrations of alpha-fetoprotein.

[0069] All ROC plots shown in Figures 1 A to 19A (in the A Figures) as well as the corresponding confusion matrices shown in confusion matrices illustrating Figures 1 B to 19B (in the B Figures) were obtained by analyzing blood serum samples of individuals with or without hepatocellular carcinoma. The recruiting of patients for the analyzed samples as well as the sample preparation and measuring will be explained in the following in more detail.

[0070] Study design

[0071] A retrospective case-control design using banked serum samples from individuals with or without HCC was chosen. The samples were either banked or retained samples from routine clinical care collected for other research studies or non-research purposes, or samples collected as part of a clinical study. Serum cohorts were obtained from seven different study sites, namely five in Europe and two in the U.S. Thus, this study was a retrospective, multicentric, cross-sectional case-control study.

[0072] Group assignment and reference standard

[0073] The case group comprised patients with diagnosis of primary, early-stage HCC as defined by Milan criteria. Patients with late-stage HCC (beyond Milan criteria) were defined as additional group. Reference standard for HCC diagnosis was high-resolution imaging (contrast-enhanced computed tomography (CT) or magnetic resonance imaging (MRI)) and / or biopsy according to current clinical guidelines. The control group were patients without clinical signs of HCC, i.e., negative surveillance tests (abdominal ultrasound and / or AFP).

[0074] Eligibility criteria

[0075] Serum samples had to comply with the following inclusion criteria to be eligible for the study:

[0076] • Serum collected from patient with or without primary HCC

[0077] • Sampling at the time (±2 months) of HCC diagnosis (MRI, CT or biopsy), or of imaging showing no HCC for control group

[0078] • Samples stored at -80°C • Storage time <5 years

[0079] • Maximum 1 freeze / thaw cycle

[0080] • Samples with AFP determination within up to three months (before and after) blood withdrawal • Serum from patients of age >18 years

[0081] However, samples were excluded if the patients have already been treated for the given HCC lesion at the time of the sample collection. Cohort details

[0082] Details of the tested patient cohort are listed in the following Table A:

[0083] Table A: Details of the tested patient cohort.

[0084] Characteristic N Train, N = 641 Test, N = 155

[0085] Center, n / N (%) 796

[0086] Center 1a 169 / 641 (26%) 0 / 155(0%)

[0087] Center 1b 34 / 641 (5.3%) 0 / 155(0%)

[0088] Center 2 122 / 641 (19%) 0 / 155(0%)

[0089] Center 3 35 / 641 (5.5%) 0 / 155 (0%)

[0090] Center 4 78 / 641 (12%) 0 / 155 (0%)

[0091] Centers 165 / 641 (26%) 155 / 155(100%)

[0092] Center 6 38 / 641 (5.9%) 0 / 155 (0%)

[0093] Age, Mean (SD) 796 60.07 (9.85) 58.00 (10.09)

[0094] Age Group, n / N (%) 796

[0095] <63 356 / 641 (56%) 104 / 155 (67%)

[0096] >=63 285 / 641 (44%) 51 / 155 (33%)

[0097] Sex, n / N (%) 796 female 193 / 641 (30%) 62 / 155 (40%) male 448 / 641 (70%) 93 / 155 (60%)

[0098] HCC Stage (Milan)*, n / N (%) 796 early 172 / 641 (27%) 39 / 155 (25%) negative 469 / 641 (73%) 116 / 155 (75%)

[0099] AFP, Median (Mean) 796 5.00 (10.79) 5.08(19.41) Characteristic N Train, N = 641 Test, N = 155

[0100] BCLC-Stage, n / N (%) 796 negative 469 / 641 (73%) 116 / 155 (75%)

[0101] 0 29 / 641 (4.5%) 3 / 155 (1.9%)

[0102] A 121 / 641 (19%) 32 / 155 (21 %)

[0103] B 1 / 641 (0.2%) 0 / 155 (0%)

[0104] C 14 / 641 (2.2%) 0 / 155(0%)

[0105] D* 7 / 641 (1.1%) 0 / 155(0%)

[0106] D 0 / 641 (0%) 4 / 155 (2.6%)

[0107] Child-Pugh Score, n / N (%) 763

[0108] 5 168 / 608 (28%) 49 / 155 (32%)

[0109] 6 108 / 608(18%) 38 / 155 (25%)

[0110] 7 95 / 608 (16%) 33 / 155 (21 %)

[0111] 8 74 / 608(12%) 11 / 155(7.1%)

[0112] 9 56 / 608 (9.2%) 11 / 155 (7.1%)

[0113] 10 44 / 608 (7.2%) 9 / 155 (5.8%)

[0114] 11 29 / 608(4.8%) 2 / 155(1.3%)

[0115] 12 27 / 608(4.4%) 2 / 155(1.3%)

[0116] 13 7 / 608(1.2%) 0 / 155(0%)

[0117] (Missing) 33 0

[0118] Child-Pugh Class, n / N (%) 792

[0119] Class A 299 / 637 (47%) 87 / 155 (56%)

[0120] Class B 227 / 637 (36%) 55 / 155 (35%)

[0121] Class C 111 / 637(17%) 13 / 155(8.4%)

[0122] (Missing) 4 0

[0123] * The target variable used in modelling for this use case. Milan criteria stratifies HCC into early-stage disease (transplant-appropriate, curative intent) and late-stage (transplant inappropriate) based on lesion size (one lesion smaller than 5 cm; alternatively, up to three lesions, each smaller than 3 cm), no extrahepatic manifestations and no gross vascular invasion.

[0124] Sample preparation

[0125] Samples were prepared with reagents of the AXINON® serum kit 2.0 offered by numares AG.

[0126] During sample preparation, all reagents were used at ambient temperature (15-30°C). a) Calibration samples

[0127] The AXINON® serum calibrator 2.0 was filled into an NMR tube. A cap for NMR tubes was placed onto the NMR tube. Subsequently, the calibration sample was placed in a defined position of an NMR rack. b) Control samples

[0128] The AXINON® serum control 2.0 was filled into an NMR tube. A cap for NMR tubes was placed onto the NMR tube. Subsequently, two control samples were placed in defined positions of the same NMR rack. c) Analytical samples

[0129] The AXINON® serum additives solution 2.0 was combined with the blood serum to be analyzed in a ratio of 1 :10 in a suitable reagent container. The liquid was gently mixed, wherein foaming was avoided. The volume of the mixture required for the NMR measurement was transferred into an NMR tube. A cap for NMR tubes was placed onto the NMR tube. Each analytical sample was placed at a defined position into the same NMR rack according to a rack list (as defined by the AXINON® Sample Wizard, see AXINON® Sample Wizard User manual). Attention was paid to ensure proper positioning of the samples. The analysis results of samples that were not clearly assignable were discarded.

[0130] Measurement

[0131] All measurements were carried out on a Broker Avance II+ 600MHz NMR spectrometer with an UltraShield 600 Plus NMR Magnet System, or a Broker Avance III HD 600MHz NMR spectrometer with an UltraShield 600 Plus NMR Magnet System, or a Broker Avance III HD 600MHz NMR spectrometer with an Ascend NMR Magnet System using a PATXI 1 H / D- 13C / 15N Z-GRD probe. All samples were kept at 5-7°C in the SampleJet and brought to the target temperature in the integrated preheating block before measurement.

[0132] Two measurement sequences were used for all samples:

[0133] • on the one hand, a standard pulse program with 30-degree excitation pulse and presaturation for water suppression was used (zgpr30);

[0134] • on the other hand, a pulse program to measure T2-weighted spectra without J modulation using refocusing pulses between double spin echoes (project = Periodic Refocusing Of J Evolution by Coherence Transfer) was used.

[0135] Samples were measured in batches of up to 93 analytical samples per run. In addition to the analytical samples, each run included one AXINON® blood serum calibrator sample and two AXINON® blood serum control samples (before and after the analytical blood serum samples, respectively) to assure ideal measurement conditions throughout the run.

[0136] Signal analysis a) Spectrum Qualification (Quality Control for measurement)

[0137] NMR spectra underwent automatic referencing, phase correction and baseline correction before further analysis.

[0138] Subsequently, the NMR spectra underwent an automatic standardization and calibration procedure to minimize between-device, between-day and between-run effects. The quality of each of these spectra was assessed by a custom spectrum qualification algorithm that analyzes general spectral properties, e.g., offset and tilt of the baseline in selected spectral regions, and properties of selected indicator signals, e.g., signal position, shape and width. Spectra that did not meet the predefined quality criteria were excluded from further analysis. b) Bins

[0139] Successfully qualified spectra (typically covering a chemical shift from -5 to 14 ppm) were subjected to further modifications. In particular, broad background signals were separated with a suitable algorithm, e.g., background intensities (such as generated from proteins) were subtracted from the spectra, resulting in spectral intensities devoid of such background signals.

[0140] The cohort (i.e., the plurality) of modified spectra was checked for regions in which the cohort did not show a significant number of signals. These regions - like the region of the water signal and the regions in which signals are contained that originate from substances contained in AXINON® serum additives - were ignored in the steps explained in the following.

[0141] The remaining spectral regions were subject to an adaptive binning, which divides the spectrum in bins of differing size or extent (typically covering 0.01 to 0.05 ppm, but in extreme cases also covering 0.005 to 0.5 ppm). The resulting boundaries for the bins are tailored to represent signals and / or signal structures in the spectrum as good as possible.

[0142] Depending on the cohort of modified spectra, the size and thus the number of bins varies. Typical numbers of bins lie in a range of from 100 to 400. c) Quantifier

[0143] Quantification of substances was done by fitting a predefined, characteristic set of PseudoVoigt functions, which represent a linear combination of a Gaussian and a Lorentzian function, to the substance specific signal structure(s). The resulting signal fits were checked for goodness of fit and physical plausibility of properties related to fit parameters in order to reject results of insufficient fit quality.

[0144] Alternatively, quantification models making use of the previously assigned bins were applied. After substance identification, substance labels have been assigned to these bins. The quantification was then determined by the bin value, which calculates as [(sum of intensities in bin) / (number of data points in bin)]. The standardization by data points is used to compensate for a varying number of data points in the bins. The number of data points in a bin may vary by one data point due to shifts of the applied discretization grid.

[0145] In either case, the resulting integral of the quantification is then translated into a substance concentration by applying a conversion factor which has been determined experimentally for each of the substances.

[0146] Performance of AFP

[0147] Since alpha-fetoprotein (AFP) is the current “gold standard” biomarker used for diagnosing HCC, a comparison between the sensitivity and specificity values of the tested biomarker sets and the sensitivity and specificity of AFP was drawn. For this purpose, AFP quantification was performed with differing methods depending on availability of methods in the respective study center or biobank. AFP is a clinical standard marker which can be reliably measured with a variety of assays, e.g., with the following assays:

[0148] • Diazyme’s a-Fetoprotein Assay is based on a latex-enhanced immunoturbidimetric assay (like Cystatin C).

[0149] • Roche’s Elecsys AFP is an immunoassay for the in vitro quantitative determination of cd fetoprotein in human serum and plasma.

[0150] • The Siemens ADVIA Centaur® AFP is intended to be used for the qualitative and / or quantitative detection of alpha-fetoprotein in a clinical specimen, using a chemiluminescent immunoassay method.

[0151] The clinical use of AFP is mostly based on a fixed cutoff of AFP >= 20 ng / mL. However, showing a ROC plot based on this cutoff is not feasible. The main reason for this is that the ROC plot always shows the sensitivity and specificity for all potential cutoffs, not just for a single cutoff. Figure 19A shows the ROC plot for AFP considering all potential cutoffs. The inventors built a logistic regression model on AFP to compare the ROC plot of AFP with the ROC plots of the tested biomarker sets. For the Tables shown in the following sections, an AFP cutoff of 20 ng / mL was used to calculate the sensitivity and specificity for this given cutoff. The obtained values were compared with the corresponding values of the tested biomarker sets.

[0152] Test of identified marker substances

[0153] The identified marker substances were tested in different combinations to assess their suitability for determining the risk of an individual of having hepatocellular carcinoma. In doing so, the result of the determination based on the marker substances (predicted risk) has been checked against clinical signs of hepatocellular carcinoma in the patient who donated the serum sample, as already explained above.

[0154] The results are summarized in receiver operating characteristic (ROC) plots. In these plots, the area under the curve (AUC) indicates the fitness of the prediction. If the AUC is 0.5, the prediction is to be considered random and thus not well suited. The higher the AUC, the better is the prediction model.

[0155] However, the AUC value is not the only measure for identifying the suitability of a specific marker combination. Rather, overall values of sensitivity and specificity give additional insight in the suitability of a specific marker combination. Here, a comparison with the current “gold standard” biomarker alpha-fetoprotein (AFP) was drawn. Unexpectedly, the tested biomarker combinations turned out to have a higher sensitivity than a test based on AFP. Expressed in other words, the biomarkers, the use and application of which is subject of the present disclosure, are particularly well suited to identify a lower number of false negative individuals (patients suffering from HCC but being diagnosed as healthy) and, consequently, a higher number of true positive individuals (patients suffering from HCC and being identified as having HCC). Thus, the novel biomarkers will help in identifying more patients already having developed a hepatocellular carcinoma that would be uncovered when applying the classic analysis approaches. These patients can be subjected much earlier to an anti-HCC treatment that will ameliorate their chances of survival or even recovery.

[0156] The obtained results will be explained in the following in more detail making reference to Figures 1 A to 18B. Figures 1 A, 2A, 3A, 4A, 5A, 6A, 7A, 8A, 9A, 10A, 1 1 A, 12A, 13A, 14A, 15A, 16A, 17A, and 18A show ROC plots of different marker combinations. Figures 1 B, 2B, 3B, 4B, 5B, 6B, 7B, 8B, 9B, 10B, 1 1 B, 12B, 13B, 14B, 15B, 16B, 17B, and 18B show the corresponding confusion matrices based on the scores assigned to the patient’s health status (i.e., having HCC or not having HCC) confirmed by other tests. Figure 19A shows an ROC plot of AFP (cutoff-independent), and Figure 19B shows a confusion matrix based on the correspondingly calculated score.

[0157] The following Tables will allow additional insight into the sensitivity and specificity values of the individual tested biomarker sets.

[0158] Table 1 : Sensitivity and specificity value of a marker set comprising glutamine and proline in comparison with corresponding values of AFP. Confer Figure 1 A for the corresponding ROC plot and Figure 1 B for the corresponding confusion matrix.

[0159] DS: dataset; FN: false negative; FP: false positive; TN: true negative; TP: true positive

[0160] Table 2: Sensitivity and specificity value of a marker set comprising glycine and proline in comparison with corresponding values of AFP. Confer Figure 2A for the corresponding ROC plot and Figure 2B for the corresponding confusion matrix.

[0161] DS: dataset; FN: false negative; FP: false positive; TN: true negative; TP: true positive

[0162] Table 3: Sensitivity and specificity value of a marker set comprising glycine and valine in comparison with corresponding values of AFP. Confer Figure 3A for the corresponding ROC plot and Figure 3B for the corresponding confusion matrix.

[0163] DS: dataset; FN: false negative; FP: false positive; TN: true negative; TP: true positive

[0164] Table 4: Sensitivity and specificity value of a marker set comprising glutamine and pyruvate in comparison with corresponding values of AFP. Confer Figure 4A for the corresponding ROC plot and Figure 4B for the corresponding confusion matrix.

[0165] DS: dataset; FN: false negative; FP: false positive; TN: true negative; TP: true positive

[0166] Table 5: Sensitivity and specificity value of a marker set comprising glutamine and glycine in comparison with corresponding values of AFP. Confer Figure 5A for the corresponding ROC plot and Figure 5B for the corresponding confusion matrix.

[0167] DS: dataset; FN: false negative; FP: false positive; TN: true negative; TP: true positive

[0168] Table 6: Sensitivity and specificity value of a marker set comprising glutamine and formic acid in comparison with corresponding values of AFP. Confer Figure 6A for the corresponding ROC plot and Figure 6B for the corresponding confusion matrix.

[0169] DS: dataset; FN: false negative; FP: false positive; TN: true negative; TP: true positive Table 7: Sensitivity and specificity value of a marker set comprising alanine and formic acid in comparison with corresponding values of AFP. Confer Figure 7A for the corresponding ROC plot and Figure 7B for the corresponding confusion matrix.

[0170] DS: dataset; FN: false negative; FP: false positive; TN: true negative; TP: true positive

[0171] Table 8: Sensitivity and specificity value of a marker set comprising hydroxybutyrate and formic acid in comparison with corresponding values of AFP. Confer Figure 8A for the corresponding ROC plot and Figure 8B for the corresponding confusion matrix.

[0172] DS: dataset; FN: false negative; FP: false positive; TN: true negative; TP: true positive

[0173] Table 9: Sensitivity and specificity value of a marker set comprising hydroxybutyrate and pyruvate in comparison with corresponding values of AFP. Confer Figure 9A for the corresponding ROC plot and Figure 9B for the corresponding confusion matrix.

[0174] DS: dataset; FN: false negative; FP: false positive; TN: true negative; TP: true positive Table 10: Sensitivity and specificity value of a marker set comprising glucose and proline in comparison with corresponding values of AFP. Confer Figure 10A for the corresponding ROC plot and Figure 10B for the corresponding confusion matrix.

[0175] DS: dataset; FN: false negative; FP: false positive; TN: true negative; TP: true positive

[0176] Table 1 1 : Sensitivity and specificity value of a marker set comprising glucose and glutamine in comparison with corresponding values of AFP. Confer Figure 1 1 A for the corresponding ROC plot and Figure 1 1 B for the corresponding confusion matrix.

[0177] DS: dataset; FN: false negative; FP: false positive; TN: true negative; TP: true positive

[0178] Table 12: Sensitivity and specificity value of a marker set comprising pyruvate and lactate in comparison with corresponding values of AFP. Confer Figure 12A for the corresponding ROC plot and Figure 12B for the corresponding confusion matrix.

[0179] DS: dataset; FN: false negative; FP: false positive; TN: true negative; TP: true positive

[0180] Table 13: Sensitivity and specificity value of a marker set comprising pyruvate and HDL in comparison with corresponding values of AFP. Confer Figure 13A for the corresponding ROC plot and Figure 13B for the corresponding confusion matrix.

[0181] DS: dataset; FN: false negative; FP: false positive; TN: true negative; TP: true positive

[0182] Table 14: Sensitivity and specificity value of a marker set comprising pyruvate and HDL cholesterol in comparison with corresponding values of AFP. Confer Figure 14A for the corresponding ROC plot and Figure 14B for the corresponding confusion matrix.

[0183] DS: dataset; FN: false negative; FP: false positive; TN: true negative; TP: true positive

[0184] Table 15: Sensitivity and specificity value of a marker set comprising pyruvate and glucose in comparison with corresponding values of AFP. Confer Figure 15A for the corresponding ROC plot and Figure 15B for the corresponding confusion matrix.

[0185] DS: dataset; FN: false negative; FP: false positive; TN: true negative; TP: true positive

[0186] Table 16: Sensitivity and specificity value of a marker set comprising formic acid and lactate in comparison with corresponding values of AFP. Confer Figure 16A for the corresponding ROC plot and Figure 16B for the corresponding confusion matrix.

[0187] DS: dataset; FN: false negative; FP: false positive; TN: true negative; TP: true positive

[0188] Table 17: Sensitivity and specificity value of a marker set comprising glutamine, proline, and pyruvate in comparison with corresponding values of AFP. Confer Figure 17A for the corresponding ROC plot and Figure 17B for the corresponding confusion matrix.

[0189] DS: dataset; FN: false negative; FP: false positive; TN: true negative; TP: true positive

[0190] Table 18: Sensitivity and specificity value of a marker set comprising glutamine, glycine, and proline in comparison with corresponding values of AFP. Confer Figure 18A for the corresponding ROC plot and Figure 18B for the corresponding confusion matrix.

[0191] DS: dataset; FN: false negative; FP: false positive; TN: true negative; TP: true positive

[0192] List of references cited in the preceding sections or otherwise deemed to be relevant

[0193] 1 . Kew MC, Dos Santos HA, Sherlock S. Diagnosis of primary cancer of the liver. British medical journal. 1971 ;4(5784):408-11 .

[0194] 2. Galun D, Basaric D, Zuvela M, Bulajic P, Bogdanovic A, Bidzic N, et al. Hepatocellular carcinoma: From clinical practice to evidence-based treatment protocols. World journal of hepatology. 2015;7(20):2274-91.

[0195] 3. World Health Organization. Cancer Fact sheet N°297. [Available from: http: / / www.who.int / mediacentre / factsheets / fs297 / en / ]. 4. European Association for Study of L, European Organisation for R, Treatment of C. EASL-EORTC clinical practice guidelines: management of hepatocellular carcinoma. European journal of cancer. 2012;48(5):599-641.

[0196] 5. Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer statistics. CA: a cancer journal for clinicians. 201 1 ;61 (2):69-90.

[0197] 6. Llovet JM, A. Burroughs, and J. Bruix. Hepatocellular carcinoma. Lancet. 2003;362(9399):1907- 17.

[0198] 7. Zhang BH, Yang BH, Tang ZY. Randomized controlled trial of screening for hepatocellular carcinoma. Journal of cancer research and clinical oncology. 2004;130(7):417-22.

[0199] 8. Liver EAftSot. EASL Clinical Practice Guidelines: Management of hepatocellular carcinoma. J Hepatol. 2018.

[0200] 9. Marrero JA, et al. Diagnosis, Staging, and Management of Hepatocellular Carcinoma: 2018 Practice Guidance by the American Association for the Study of Liver Diseases. Hepatology. 2018;68(2):723-50.

[0201] 10. Omata M, et al. Asia-Pacific clinical practice guidelines on the management of hepatocellular carcinoma: a 2017 update. Hepatol Int. 2017;11 (4) :317-70.

[0202] 1 1 . Singal A, et al. Meta-analysis: surveillance with ultrasound for early-stage hepatocellular carcinoma in patients with cirrhosis. Aliment Pharmacol Ther. 2009;30(1 ):37-47.

[0203] 12. Tzartzeva K, et al. Surveillance Imaging and Alpha Fetoprotein for Early Detection of Hepatocellular Carcinoma in Patients With Cirrhosis: A Meta-analysis. Gastroenterology. 2018;154(6) :1706-18 e1 .

[0204] 13. Hoshida Y, Nijman SM, Kobayashi M, Chan JA, Brunet JP, Chiang DY, et al. Integrative transcriptome analysis reveals common molecular subclasses of human hepatocellular carcinoma. Cancer research. 2009;69(18):7385-92.

[0205] 14. Villanueva A, Minguez B, Forner A, Reig M, Llovet JM. Hepatocellular carcinoma: novel molecular approaches for diagnosis, prognosis, and therapy. Annual review of medicine. 2010 ;61 :317-28.

[0206] 15. Yamashita T, Forgues M, Wang W, Kim JW, Ye Q, Jia H, et al. EpCAM and alpha-fetoprotein expression defines novel prognostic subtypes of hepatocellular carcinoma. Cancer research. 2008;68(5):1451 -61 .

[0207] 16. Sherman M. Current status of alpha-fetoprotein testing. Gastroenterol Hepatol. 201 1 ;7(2):113-4.

[0208] 17. Lok AS, et al. Des-gamma-carboxy prothrombin and alpha-fetoprotein as biomarkers for the early detection of hepatocellular carcinoma. Gastroenterology. 2010;138(2):493-502.

[0209] 18. Marrero JA, et al. Alpha-fetoprotein, des-gamma carboxyprothrombin, and lectin-bound alphafetoprotein in early hepatocellular carcinoma. Gastroenterology. 2009;137(1 ):1 10-8.

[0210] 19. Koike Y, et al. Des-gamma-carboxy prothrombin as a useful predisposing factor for the development of portal venous invasion in patients with hepatocellular carcinoma: a prospective analysis of 227 patients. Cancer. 2001 ;91 (3):561 -9.

[0211] 20. Sterling RK, et al. Clinical utility of AFP-L3% measurement in North American patients with HCV- related cirrhosis. Am J Gastroenterol. 2007;102(10):2196-205.

[0212] 21 . Johnson PJ, et al. The detection of hepatocellular carcinoma using a prospectively developed and validated model based on serological biomarkers. Cancer Epidemiol Biomarkers Prev. 2014;23(1 ):144-53.

[0213] 22. Berhane S, et al. Role of the GALAD and BALAD-2 Serologic Models in Diagnosis of Hepatocellular Carcinoma and Prediction of Survival in Patients. . Clin Gastroenterol Hepatol. 2016;14(6):875-86 e6.

[0214] 23. Best J, et al. The GALAD scoring algorithm based on AFP, AFP-L3, and DCP significantly improves detection of BCLC early stage hepatocellular carcinoma. Z Gastroenterol. 2016;54(12):1296-305.

[0215] 24. Yang JD, et al. GALAD Score for Hepatocellular Carcinoma Detection in Comparison with Liver Ultrasound and Proposal of GALADUS Score. Cancer Epidemiol Biomarkers Prev. 2019;28(3):531 - 8. 25. Gao H, Lu Q, Liu X, Cong H, Zhao L, Wang H, et al. Application of 1 H NMR-based metabonomics in the study of metabolic profiling of human hepatocellular carcinoma and liver cirrhosis. Cancer science. 2009;100(4):782-5.

[0216] 26. Shariff Ml, Gomaa Al, Cox IJ, Patel M, Williams HR, Crossey MM, et al. Urinary metabolic biomarkers of hepatocellular carcinoma in an Egyptian population: a validation study. Journal of proteome research. 201 1 ;10(4):1828-36.

[0217] 27. Shariff Ml, Ladep NG, Cox IJ, Williams HR, Okeke E, Malu A, et al. Characterization of urinary biomarkers of hepatocellular carcinoma using magnetic resonance spectroscopy in a Nigerian population. Journal of proteome research. 2010;9(2):1096-103.

[0218] 28. Chen F, Xue J, Zhou L, Wu S, Chen Z. Identification of serum biomarkers of hepatocarcinoma through liquid chromatography / mass spectrometry-based metabonomic method. Analytical and bioanalytical chemistry. 201 1 ;401 (6) :1899-904.

[0219] 29. Chen S, Kong H, Lu X, Li Y, Yin P, Zeng Z, et al. Pseudotargeted metabolomics method and its application in serum biomarker discovery for hepatocellular carcinoma based on ultra high- performance liquid chromatography / triple quadrupole mass spectrometry. Analytical chemistry. 2013;85(17):8326-33.

[0220] 30. Chen T, Xie G, Wang X, Fan J, Qiu Y, Zheng X, et al. Serum and urine metabolite profiling reveals potential biomarkers of human hepatocellular carcinoma. Molecular & cellular proteomics : MCP. 201 1 ;10(7):M110 004945.

[0221] 31. Patterson AD, Maurhofer O, Beyoglu D, Lanz C, Krausz KW, Pabst T, et al. Aberrant lipid metabolism in hepatocellular carcinoma revealed by plasma metabolomics and lipid profiling. Cancer research. 201 1 ;71 (21 ):6590-600.

[0222] 32. Nahon P, Amathieu R, Triba MN, Bouchemal N, Nault JC, Ziol M, et al. Identification of serum proton NMR metabolomic fingerprints associated with hepatocellular carcinoma in patients with alcoholic cirrhosis. Clinical cancer research: an official journal of the American Association for Cancer Research. 2012;18(24):6714-22.

[0223] 33. Ressom HW, Xiao JF, Tuli L, Varghese RS, Zhou B, Tsai TH, et al. Utilization of metabolomics to identify serum biomarkers for hepatocellular carcinoma in patients with liver cirrhosis. Analytica chimica acta. 2012;743:90-100.

[0224] 34. Xiao JF, Varghese RS, Zhou B, Nezami Ranjbar MR, Zhao Y, Tsai TH, et al. LC-MS based serum metabolomics for identification of hepatocellular carcinoma biomarkers in Egyptian cohort. Journal of proteome research. 2012;1 1 (12):5914-23.

[0225] 35. Baniasadi H, Gowda GA, Gu H, Zeng A, Zhuang S, Skill N, et al. Targeted metabolic profiling of hepatocellular carcinoma and hepatitis C using LC-MS / MS. Electrophoresis. 2013;34(19):2910-7.

[0226] 36. Clarke JD, Novak P, Lake AD, Shipkova P, Aranibar N, Robertson D, et al. Characterization of hepatocellular carcinoma related genes and metabolites in human nonalcoholic fatty liver disease. Digestive diseases and sciences. 2014;59(2):365-74.

[0227] 37. Liu Y, Hong Z, Tan G, Dong X, Yang G, Zhao L, et al. NMR and LC / MS-based global metabolomics to identify serum biomarkers differentiating hepatocellular carcinoma from liver cirrhosis. International journal of cancer Journal international du cancer. 2014;135(3):658-68.

[0228] 38. Fitian Al, Nelson DR, Liu C, Xu Y, Ararat M, Cabrera R. Integrated metabolomic profiling of hepatocellular carcinoma in hepatitis C cirrhosis through GC / MS and UPLC / MS-MS. Liver international: official journal of the International Association for the Study of the Liver. 2014;34(9):1428-44.

[0229] 39. Zeng J, Yin P, Tan Y, Dong L, Hu C, Huang Q, et al. Metabolomics study of hepatocellular carcinoma: discovery and validation of serum potential biomarkers by using capillary electrophoresis-mass spectrometry. Journal of proteome research. 2014;13(7):3420-31 .

[0230] 40. Bowers J, Hughes E, Skill N, Maluccio M, Raftery D. Detection of hepatocellular carcinoma in hepatitis C patients: biomarker discovery by LC-MS. Journal of chromatography B, Analytical technologies in the biomedical and life sciences. 2014;966:154-62.

[0231] 41. Shao Y, Zhu B, Zheng R, Zhao X, Yin P, Lu X, et al. Development of urinary pseudotargeted LC- MS-based metabolomics method and its application in hepatocellular carcinoma biomarker discovery. Journal of proteome research. 2015;14(2):906-16. 42. Yin P, Wan D, Zhao C, Chen J, Zhao X, Wang W, et al. A metabonomic study of hepatitis B-induced liver cirrhosis and hepatocellular carcinoma by using RP-LC and HILIC coupled with mass spectrometry. Molecular bioSystems. 2009;5(8):868-76.

[0232] 43. Wu H, Xue R, Dong L, Liu T, Deng C, Zeng H, et al. Metabolomic profiling of human urine in hepatocellular carcinoma patients using gas chromatography / mass spectrometry. Analytica chimica acta. 2009;648(1 ):98-104.

[0233] 44. Wang B, Chen D, Chen Y, Hu Z, Cao M, Xie Q, et al. Metabonomic profiles discriminate hepatocellular carcinoma from liver cirrhosis by ultraperformance liquid chromatography-mass spectrometry. Journal of proteome research. 2012;1 1 (2):1217-27.

[0234] 45. Tang, Ling et al. “Global Metabolic Profiling Identifies a Pivotal Role of Proline and Hydroxyproline Metabolism in Supporting Hypoxic Response in Hepatocellular Carcinoma.” Clinical cancer research: an official journal of the American Association for Cancer Research vol. 24,2 (2018): 474- 485. doi :10.1 158 / 1078-0432. CCR-17-170

[0235] 46. D'Aniello, Cristina et al. “Proline Metabolism in Tumor Growth and Metastatic Progression.” Frontiers in oncology vol. 10 776. 15 May. 2020, doi:10.3389 / fonc.2020.00776

[0236] 47. Zhang, Taotao et al. “Mitochondrial GCN5L1 regulates glutaminase acetylation and hepatocellular carcinoma.” Clinical and translational medicine vol. 12,5 (2022): e852. doi:10.1002 / ctm2.852

[0237] 48. Li S, Gao D, Jiang Y. Function, Detection and Alteration of Acylcarnitine Metabolism in Hepatocellular Carcinoma. Metabolites. 2019; 9(2):36.

[0238] 49. Luo, Wenqi et al. “Decreased expression of 3-hydroxybutyrate dehydrogenase 1 is a prognostic marker and promotes tumor progression in hepatocellular carcinoma.” Pathology, research and practice, vol. 238 1541 1 1 . 6 Sep. 2022, doi :10.1016 / j .prp.2022.1541 1 1

[0239] 50. Gao, Hongchang et al. “Application of 1 H NMR-based metabonomics in the study of metabolic profiling of human hepatocellular carcinoma and liver cirrhosis.” Cancer science vol. 100,4 (2009): 782-5. doi :10.1 1 1 1 Zj.1349-7006.2009.01086.x

[0240] 51. Gong ZG, Zhao W, Zhang J, Wu X, Hu J, Yin GC, Xu YJ. Metabolomics and eicosanoid analysis identified serum biomarkers for distinguishing hepatocellular carcinoma from hepatitis B virus- related cirrhosis. Oncotarget. 2017 Jul 10;8(38):63890-63900. doi: 10.18632 / oncotarget.19173. PMID: 28969038; PMCID: PMC5609970.

[0241] 52. Wu, Hao et al. “Metabolomic profiling of human urine in hepatocellular carcinoma patients using gas chromatography / mass spectrometry.” Analytica chimica acta vol. 648,1 (2009): 98-104. doi : 10.1016 / j. aca.2009.06.033

[0242] 53. Jiang JT, Xu N, Zhang XY, Wu CP. Lipids changes in liver cancer. J Zhejiang Univ Sci B. 2007 Jun;8(6):398-409. doi: 10.1631 / jzus.2007.B0398.

[0243] 54. WO 2022 / 240891 A1

[0244] 55. US 2014 / 0134751 A1

[0245] 56. US 2021 / 0208147 A1

Claims

Claims1. Use of a marker set comprising at least two substances in an in vitro method for determining the risk of an individual to have hepatocellular carcinoma, characterized in that a first of the at least two substances is chosen from a first group consisting of glutamine, proline, valine, pyruvate, glycine, glutamate, hydroxybutyrate, alanine, and formic acid, and wherein a second of the at least two substances is chosen from the first group, from a second group, or from a third group, wherein the second group consists of lactate and glucose, and wherein the third group consists of high-density lipoprotein, apolipoprotein A1 , low-density lipoprotein, apolipoprotein B-48, apolipoprotein B-100, triglycerides, total cholesterol, LDL cholesterol, and HDL cholesterol.

2. Use according to claim 1 , characterized in that the second substance is different from glutamate if the first substance is glutamine, and in that the second substance is different from glutamine if the first substance is glutamate.

3. Use according to claim 1 or 2, characterized in that the first substance is glutamine and the second substance is proline.

4. Use according to claim 1 or 2, characterized in that the first substance is glycine and the second substance is proline.

5. Use according to claim 1 or 2, characterized in that the first substance is glycine and the second substance is valine.

6. Use according to claim 1 or 2, characterized in that the first substance is glutamine and the second substance is pyruvate.

7. Use according to claim 1 or 2, characterized in that the first substance is glutamine and the second substance is glycine.

8. Use according to claim 1 or 2, characterized in that the first substance is glutamine and the second substance is formic acid.

9. Use according to claim 1 or 2, characterized in that the first substance is alanine and the second substance is formic acid.

10. Use according to claim 1 or 2, characterized in that the first substance is glutamine and the second substance is glucose.

11. Use according to any of the preceding claims, characterized in that the marker set additionally comprises at least one of alpha-fetoprotein, alpha-fetoprotein isoform L3, and des-y-carboxyprothrombin.

12. Use according to any of the preceding claims, characterized in that the method comprises determining a stage of the hepatocellular carcinoma on a predeterminable scale.

13. Marker set comprising at least two substances for use in in-vivo diagnostics of hepatocellular carcinoma, characterized in that a first of the at least two substances is chosen from a first group consisting of glutamine, proline, valine, pyruvate, glycine, glutamate, hydroxybutyrate, alanine, and formic acid, and wherein a second of the at least two substances is chosen from the first group, from a second group, or from a third group, wherein the second group consists of lactate and glucose, and wherein the third group consists of high-density lipoprotein, apolipoprotein A1 , low-density lipoprotein, apolipoprotein B-48, apolipoprotein B-100, triglycerides, total cholesterol, LDL cholesterol, and HDL cholesterol.

14. Method for analyzing an isolated body fluid sample in vitro, comprising the following steps: a) determining the concentration of least two substances, wherein a first of the at least two substances is chosen from a first group consisting of glutamine, proline, valine, pyruvate, glycine, glutamate, hydroxybutyrate, alanine, and formic acid, and wherein a second of the at least two substances is chosen from the first group, from a second group, or from a third group, wherein the second group consists of lactate and glucose, and wherein the third group consists of high-density lipoprotein, apolipoprotein A1 , low- density lipoprotein, apolipoprotein B-48, apolipoprotein B-100, triglycerides, totalcholesterol, LDL cholesterol, and HDL cholesterol in an isolated body fluid sample from an individual by analyzing the body fluid sample with a suited measuring technique, b) calculating a score from the determined concentrations, the score being indicative for determining the risk of the individual to have hepatocellular carcinoma.

15. Method according to claim 14, characterized in that calculating the score involves calculating a ratio between at least two concentration values.

Images