A new model for the diagnosis of liver fibrosis in patients with metabolic dysfunction-associated steatotic liver disease

Abstract

The present invention relates to an in vitro method for assessing the presence of liver fibrosis in an individual suffering from MASLD. The method comprises analyzing concentration values of a plurality of biomarkers in a sample isolated from the individual and entering the obtained concentration values into an algorithm to generate a classification of the sample as positive or negative for liver fibrosis. The algorithm is trained with a plurality of independent training samples isolated from individuals with and without liver fibrosis, wherein the concentration values of the biomarkers were evaluated by means of logistic regression analysis.

Description

[0001] DESCRIPTION TITLE

[0002] A new model for the diagnosis of liver fibrosis in patients with metabolic dysfunction-associated steatotic liver disease

[0003] Technical field

[0004] The present invention relates to an in vitro or clinical-biochemical method for evaluating the presence of liver fibrosis in a patient with MASLD.

[0005] Prior art

[0006] Metabolic Dysfunction-Associated Steatotic Liver Disease (MASLD) is characterized by a wide spectrum of liver damage ranging from simple steatosis (accumulation of fat in the liver) to necro-inflammation with or without fibrosis (Metabolic Dysfunction-Associated Steatohepatitis, MASH), which in turn can progress to liver cirrhosis and promote the onset of complications related thereto such as hepatocellular carcinoma. Recent studies show that the prevalence of MASLD in the general population varies between 25% and 30%, rising to 40-70% in patients with type 2 diabetes mellitus (T2DM) and 60-95% in overweight / obese subjects.

[0007] Therefore, MASLD has become, in recent years, the chronic liver disease that in clinical practice raises the most concerns in view of its potential progression to more advanced liver disease and the relative complications. In fact, MASH lends itself to becoming the main indication for liver transplantation within a decade. Furthermore, growing clinical and epidemiological evidence shows that MASLD is associated not only with the development of liver events but also with an increased risk of cardiovascular events.

[0008] The onset and progression of liver damage in MASLD is closely associated with the presence of metabolic alterations such as insulin resistance, overweight / obesity and T2DM. Among patients with MASLD, those with MASH have a double risk of progression to fibrosis, which is the main driver towards progression into cirrhosis and its complications, including HCC. Therefore, an accurate stratification of fibrotic risk is crucial to identify at-risk patients on which to intervene to stop / delay the progression of damage. The gold standard for this purpose is liver biopsy, an expensive, invasive examination that is not without risks for the patient. Currently, the non-invasive method used to identify patients with more advanced liver fibrosis is Fibroscan, whose accuracy reaches AUC values ranging from 0.82 to 0.88; however, this procedure requires the use of expensive instruments, present only in reference institutions for the diagnosis and management of the disease. Other non-invasive tools based on anthropometric, clinical and biochemical parameters are available (Fibrosis-4, FIB-4; NAFLD Fibrosis Score, NFS; Enhanced Liver Fibrosis, ELF test), however these instruments have a much lower accuracy with respect to Fibroscan and in any case are not all easily applicable on a large scale.

[0009] Therefore, it would be important to identify and validate new non-invasive biomarkers or scores that can be used for risk stratification and / or monitoring the progression of MASLD and liver fibrosis, without resorting to liver biopsy.

[0010] Summary of the invention

[0011] A first aspect of the present invention relates to an in vitro, i.e. clinical- biochemical, method for assessing the presence of liver fibrosis in an individual comprising the steps of: a) measuring in a blood sample of the individual the concentration values of the following biomarkers: cortisol, cortisone, corticosterone, testosterone, androsterone glucuronide, androsterone sulphate, etiocholanolone glucuronide, etiocholanolone sulphate, 5a-androstan- 3a, 17|3-diol 3-glucuronide and 5[3-androstan-3a,17[3-diol 3-glucuronide. b) analyzing the values obtained in step a) by means of an algorithm trained to classify said sample as positive or negative for liver fibrosis, wherein the algorithm is a logistic regression model in which each concentration value obtained in step a) is associated with a coefficient obtained by training the algorithm.

[0012] Preferably, the training of the algorithm is obtained by processing some values of the at least six biomarkers obtained from a plurality of independent training samples from said sample, said training samples being isolated from individuals having liver fibrosis and individuals not having liver fibrosis.

[0013] Preferably, the individual is suffering from hepatic steatosis, more preferably he / she is suffering from steatohepatitis associated with liver fibrosis.

[0014] Preferably, the method further comprises 1 ) a step of evaluating clinical data of the individual selected from: age, Body Mass Index (BMI) and the presence or absence of type 2 diabetes mellitus and 2) the dosage of simple and repeatable blood values (haemoglobin and platelets from blood count, ALT and albumin from a serum sample).

[0015] Preferably, the logistic regression model comprises a constant, preferably comprised between 7 and 10.

[0016] A second aspect of the present invention relates to a system for assessing a liver disease in an individual. Preferably, the system comprises: a liquid chromatography and tandem mass spectrometry (LC-MS / MS) analyzer configured to measure the levels of a plurality of steroid biomarkers in a sample isolated from an individual, wherein the plurality of steroid biomarkers comprises the following biomarkers: cortisol, cortisone, corticosterone, testosterone, androsterone glucuronide, androsterone sulphate, etiocholanolone glucuronide, etiocholanolone sulphate, 5a- androstan-3a ,17|3-diol 3-glucuronide and 5[3-androstan-3a , 17|3-diol 3- glucuronide; and a processor capable of executing an algorithm trained with a plurality of independent training samples, wherein said algorithm is configured to generate a classification of said sample as positive or negative for liver fibrosis based on the measured levels of the plurality of steroid biomarkers.

[0017] A third aspect of the present invention relates to a non-transitory computer-readable medium containing instructions suitable for inducing the system described above to perform the method for assessing liver fibrosis in an individual.

[0018] Detailed description of preferred embodiments of the invention

[0019] A first aspect of the present invention relates to an in vitro method for assessing the presence of a liver disease in an individual. The method involves the use of an extended panel of steroid biomarkers, which are analyzed in a biological sample isolated from the individual. The levels of these biomarkers are then fed into an algorithm that has been trained with a plurality of independent training samples. The algorithm generates a classification of the sample as positive or negative for the presence of liver disease, thus allowing to assess the presence or absence of a liver disease in the individual.

[0020] In an embodiment, the method includes analyzing the levels of a plurality of steroid biomarkers in a sample isolated from the individual.

[0021] In a preferred embodiment of the invention, the liver disease is a liver fibrosis, preferably, an advanced liver fibrosis.

[0022] In an embodiment, the individual is suffering from a hepatic steatosis, preferably selected from: metabolic dysfunction-associated steatotic liver disease (MASLD) and metabolic-dysfunction-associated steatohepatitis (MASH). More preferably, the individual is suffering from MASLD.

[0023] Preferably, the isolated sample is selected from blood, serum or plasma, or any other suitable biological sample. More preferably, the isolated sample is a blood sample. Preferably, the plurality of steroid biomarkers comprises or consists of: cortisol, cortisone, corticosterone, testosterone, androsterone glucuronide, androsterone sulphate, etiocholanolone glucuronide, etiocholanolone sulphate, 5a-androstan-3a,17[3-diol 3- glucuronide and 5[3-androstan-3a,17[3-diol 3-glucuronide. In some cases, all ten of these biomarkers can be analyzed, while in other cases only a subset of these biomarkers can be analyzed.

[0024] Preferably, after determining the levels of the steroid biomarkers, said levels are fed to an algorithm. In an embodiment, the algorithm is trained with a plurality of independent training samples, which may include samples from individuals suffering from a liver disease and samples from individuals not suffering from a liver disease.

[0025] The algorithm can be a logistic regression model or any other type of machine learning or statistical model.

[0026] Preferably, the algorithm generates a classification of the sample as positive or negative for a liver disease based on the measured levels of the steroid biomarkers. This classification can be used to determine the presence or absence of a liver disease in the individual. In some cases, the algorithm can also use additional clinical data of the individual, such as age, body mass index (BMI), and presence or absence of type 2 diabetes mellitus, to generate the classification. The algorithm can also use the dosage of other blood values, such as haemoglobin, platelets, ALT, and albumin.

[0027] In an embodiment, the method further comprises a step of determining the presence or absence of liver disease in the individual based on the classification generated by the algorithm. This determination can be used to guide further diagnostic or therapeutic actions. For example, if the classification indicates the presence of liver disease, the individual may be directed to further tests or treatments. Conversely, if the classification indicates the absence of advanced liver fibrosis, the individual is spared unnecessary invasive procedures such as liver biopsy.

[0028] In an embodiment, the method comprises analyzing a plurality of steroid biomarkers in the sample.

[0029] Said plurality of biomarkers is selected from the group consisting of: cortisone, cortisol, 21 -deoxycortisol, corticosterone, 11 -deoxycortisol, androstenedione, 11 -deoxycorticosterone, testosterone, dehydroepiandrosterone, 17a-hydroxyprogesterone, epitestosterone, dihydrotestosterone, progesterone, 5a-androstan-3a 17|3-diol 3- glucuronide, 5[3-androstan-3a 17|3-diol 3-glucuronide, 5a-androstan-3a 17|3-diol 17-glucuronide, 5[3-androstan-3a 17|3-diol 17-glucuronide, etiocholanolone glucuronide, androsterone glucuronide, testosterone glucuronide, dehydroepiandrosterone sulphate, testosterone sulphate, epitestosterone sulphate, epiandrosterone sulphate, etiocholanolone sulphate and androsterone sulphate.

[0030] Preferably, said steroid biomarkers are at least six biomarkers selected from a group consisting of cortisol, cortisone, corticosterone, testosterone, androsterone glucuronide, androsterone sulphate, etiocholanolone glucuronide, etiocholanolone sulphate, 5a-androstan-3a,17[3-diol 3- glucuronide and 5[3-androstan-3a,17[3-diol 3-glucuronide. Preferably, the method includes the analysis of at least seven, at least eight, more preferably at least nine or of the ten steroid biomarkers mentioned above. In an embodiment, the method can include the analysis of the levels of these steroid biomarkers using a technology known to the person skilled in the art, such as for example liquid chromatography coupled with mass spectrometry (LC-MS / MS). This technique can provide accurate and reliable measurements of the levels of these biomarkers in the sample. The use of this advanced technology can improve the accuracy and reliability of the liver disease assessment.

[0031] In an embodiment, the method comprises the use of an algorithm to generate a classification of the sample as positive or negative for liver disease. The algorithm can be a logistic regression model, a type of statistical model commonly used for binary classification problems. The logistic regression model can be trained with a plurality of independent training samples, which can include samples from individuals with known liver disease status. The training process can involve adjusting the model parameters to minimize the difference between the model predictions and the actual results in the training samples.

[0032] In an embodiment, the logistic regression model uses specific coefficients for each input variable. These coefficients are preferably determined during the training process and reflect the relative importance of each input variable in predicting the outcome. For example, the coefficient for cortisone may be different from that for androsterone sulphate, reflecting the different role that these biomarkers can play in liver disease.

[0033] In an embodiment, the logistic regression model comprises a constant, preferably comprised between 7 and 10. This constant, also known as an intercept, can be added to the sum of the products of the input variables and their coefficients. The constant term is preferably determined during the training process and may represent the baseline risk of developing liver disease in the absence of other information.

[0034] In an embodiment, the liver disease assessment method also envisages obtaining clinical data and further biochemical data from the individual. This clinical data can provide additional information that can be used to improve the accuracy of the liver fibrosis assessment. Preferably, the clinical data comprise age, body mass index (BMI), and the presence or absence of type 2 diabetes mellitus. The further biochemical data are some blood values, such as haemoglobin, platelets, ALT and albumin.

[0035] In an embodiment, the age of the individual is used as an input variable in the algorithm. Age can be a significant factor in the development of liver disease, as the risk of liver disease can increase with age. The age of the individual is expressed in years with a chosen cut-off greater than or equal to 50 years. The presence of age greater than or equal to 50 years is encoded as 1 and the age less than 50 years is encoded as 0. The age coefficient greater than or equal to 50 years is comprised between 0.2 and 1 , preferably between 0.4 and 0.8.

[0036] In an embodiment, the body mass index (BMI) of the individual is used as an input variable in the algorithm. BMI, which is a measure of body fat based on height and weight, may be associated with the risk of liver disease. For example, individuals with a higher BMI may have a higher risk of liver disease due to the association between obesity and liver disease. The individual's BMI can be calculated using standard formulas and can be expressed in units of kg / m2. The BMI coefficient is comprised between 0.04 and 0.12, preferably between 0.06 and 0.1.

[0037] In an embodiment, the presence or absence of type 2 diabetes mellitus may be used as an input variable in the algorithm. Type 2 diabetes mellitus may be associated with the risk of liver disease, as individuals with type 2 diabetes mellitus may have an increased risk of liver disease. The presence or absence of type 2 diabetes mellitus can be determined by standard diagnostic methods and can be encoded as a binary variable, with the presence of type 2 diabetes mellitus encoded as 1 and the absence of type 2 diabetes mellitus encoded as 0. The coefficient of type 2 diabetes mellitus is comprised between 0.7 and 1 .6, preferably between 0.9 and 1 .4.

[0038] In an embodiment, the alanine aminotransferase (ALT) of the individual is used as an input variable in the algorithm. ALT is measured on the patient's serum sample and can be considered a measure of liver health. For example, individuals with elevated ALT may have an increased risk of liver disease. ALT is measured in IU / L. The ALT coefficient is comprised between 0.001 and 0.2, preferably between 0.005 and 0.1 .

[0039] In an embodiment, the haemoglobin of the individual is used as an input variable in the algorithm. Haemoglobin is measured on the patient's blood sample and is related to liver health. For example, individuals with liver diseases have lower levels of haemoglobin. Haemoglobin is measured in g / dL. The haemoglobin coefficient is comprised between 0.01 and 1 , preferably between 0.1 and 0.7.

[0040] In an embodiment, the albumin of the individual is used as an input variable in the algorithm. Albumin is measured on the patient's serum sample and its value is related to liver health. For example, individuals with liver diseases have lower levels of albumin. Albumin is measured in g / dL. The albumin coefficient is comprised between -1.2 and -0.001 , preferably between -1 .0 and -0.2.

[0041] In an embodiment, the platelets of the individual are used as an input variable in the algorithm. Platelets are measured on a patient's blood sample and their value is related to liver health. For example, individuals with liver diseases have lower levels of platelets. The platelets are counted on a volume of blood (value * 109 / L). The platelet coefficient is comprised between -0.5 and 0.5, preferably between -0.3 and 0.3.

[0042] In an embodiment, the cortisol is expressed as the base 10 logarithm of the nanograms per milliliter (ng / mL) of cortisol detected in the isolated sample. The cortisol coefficient is comprised between 2.3 and 2.8, preferably between 2.45 and 2.65.

[0043] In an embodiment, the cortisone is expressed as the base 10 logarithm of the nanograms per milliliter (ng / mL) of cortisone detected in the isolated sample. The cortisone coefficient is comprised between -2.3 and -2.8, preferably between -2.45 and -2.65.

[0044] In an embodiment, the corticosterone is expressed as the base 10 logarithm of the picograms per milliliter (ng / mL) of corticosterone detected in the isolated sample. The corticosterone coefficient is comprised between -1 .0 and -1 .7, preferably between -1 .2 and -1 .5.

[0045] In an embodiment, testosterone is expressed as the base 10 logarithm of the picograms per milliliter (pg / mL) of testosterone detected in the isolated sample. The testosterone coefficient is comprised between 0.4 and 1.1 , preferably between 0.6 and 0.9.

[0046] In an embodiment, the androsterone glucuronide is expressed as the base 10 logarithm of the pg / mL of androsterone glucuronide detected in the isolated sample. The androsterone glucuronide coefficient is comprised between 2.9 and 4, preferably between 3.2 and 3.7.

[0047] In an embodiment, the androsterone sulphate is expressed as the base 10 logarithm of the ng / mL of androsterone sulphate detected in the isolated sample. The androsterone sulphate coefficient is comprised between -1.8 and -2.9, preferably between -2 and -2.7.

[0048] In an embodiment, the etiocholanolone glucuronide is expressed as the base 10 logarithm of the pg / mL of etiocholanolone glucuronide detected in the isolated sample. The coefficient of the etiocholanolone glucuronide is comprised between -1 .2 and -2.3, preferably between -1 .4 and -2.1 .

[0049] In an embodiment, the etiocholanolone sulphate is expressed as the base 10 logarithm of the picograms per milliliter (ng / mL) of etiocholanolone sulphate detected in the isolated sample. The coefficient of the etiocholanolone sulphate is comprised between 1.3 and 1.9, preferably between 1 .4 and 1 .6.

[0050] In an embodiment, the 5a-androstan-3a , 17|3-diol 3-glucuronide is expressed as the base 10 logarithm of the pg / mL of 5a-androstan-3a , 17(3- diol 3-glucuronide detected in the isolated sample. The coefficient of the 5a-androstan-3a,17[3-diol 3-glucuronide is comprised between -0.7 and - 1 .6, preferably between -0.9 and -1 .4.

[0051] In an embodiment, the 5a-androstan-3a ,17|3-diol 17-glucuronide is expressed as the base 10 logarithm of the pg / mL of 5a-androstan-3a ,17(3- diol 17-glucuronide detected in the isolated sample. The coefficient of the 5a-androstan-3a,17[3-diol 17-glucuronide is comprised between -0.8 and - 1 .83, preferably between -1 .1 and -1 .53.

[0052] In an embodiment, the algorithm can use the clinical data and the generic blood values, in addition to the measured levels of the steroid biomarkers, to generate the classification of the sample as positive or negative for liver disease. The inclusion of clinical data and generic blood variables in the algorithm may improve the accuracy of liver fibrosis assessment by taking into account factors that may influence the development and progression of liver fibrosis.

[0053] In an embodiment, the algorithm generates a classification of the sample as positive or negative for liver disease, preferably for MASLD, more preferably for MASLD with liver fibrosis based on specific cutoff values for a calculated value, called Y. The Y value can be derived from the developed algorithm, which can incorporate measured levels of steroid biomarkers and clinical data. The specific cutoff values for Y can be determined based on the training data and can be optimized to maximize the accuracy of the liver fibrosis assessment.

[0054] In a preferred embodiment of the invention, the sample is positive for advanced liver fibrosis if the Y value is equal to or greater than 2, preferably greater than or equal to 2.72. The sensitivity and specificity of this classification are 16.5% and 100%, respectively.

[0055] In a preferred embodiment of the invention, the sample is negative for liver fibrosis if the Y value is less than -3, more preferably if it is less than -3.39. This classification may indicate a low probability of advanced liver fibrosis in the individual. The sensitivity and specificity of this classification can be 100% and 17.5%, respectively.

[0056] In a preferred embodiment of the invention, the algorithm classifies the sample as diagnostic doubt if the Y value is comprised between -3 and 2, preferably if it is comprised between -3.39 and 2.72. This classification may indicate an uncertainty about the presence of advanced liver fibrosis in the individual. In these cases, at the discretion of the doctor, a liver biopsy may be indicated to confirm the diagnosis.

[0057] In an embodiment, the algorithm is implemented by a processor.

[0058] Advantageously, the method can potentially prevent liver biopsy in 81% of patients. This result can be achieved by accurately classifying most samples as positive or negative for liver fibrosis, thus reducing the need for invasive diagnostic procedures. This potential advantage can make the method a valuable tool in the management of liver diseases.

[0059] A second aspect of the present invention relates to a system for assessing the presence of a liver disease in an individual. The system comprises a liquid chromatography analyzer, preferably an LC-MS / MS and a processor. Preferably, the liquid chromatography analyzer is configured to measure the levels of a plurality of steroid biomarkers in a sample isolated from an individual.

[0060] In an embodiment, the plurality of steroid biomarkers comprises at least six biomarkers selected from the group consisting of cortisol, cortisone, corticosterone, testosterone, androsterone glucuronide, androsterone sulphate, etiocholanolone glucuronide, etiocholanolone sulphate, 5a- androstan-3a,17[3-diol 3-glucuronide and 5[3-androstan-3a,17[3-diol 3- glucuronide. In some cases, the liquid chromatography analyzer is configured to measure the levels of all ten of these biomarkers.

[0061] The processor of the system is preferably configured to execute an algorithm. This algorithm can be trained with a plurality of independent training samples. Preferably, the algorithm is configured to generate a classification of the sample as positive or negative for a liver disease based on the measured levels of the plurality of steroid biomarkers. In some cases, the algorithm executed by the processor may be a logistic regression model. In a preferred embodiment of the invention, the liver disease is a hepatic steatosis, preferably selected from: metabolic dysfunction-associated steatotic liver disease (MASLD) and metabolic- dysfunction-associated steatohepatitis (MASH). More preferably, the liver disease is MASLD.

[0062] In a preferred embodiment, the algorithm is as described in detail above.

[0063] In an embodiment, the system comprises at least one input interface. This input interface can be configured to receive the individual's clinical data and the values of some other simple biochemical tests. Preferably, the clinical data comprise at least one of the following elements: age, body mass index (BMI), and the presence or absence of type 2 diabetes mellitus. The algorithm executed by the processor is preferably configured to use the clinical data to generate the test result. The algorithm executed by the processor can be further configured to use the clinical data to generate the classification of the sample as positive or negative for liver fibrosis. Preferably, the simple biochemical data comprises at least one of the following elements: haemoglobin, platelets, ALT and albumin. The algorithm executed by the processor is preferably configured to use the simple biochemical data to generate the test result. The algorithm executed by the processor can be further configured to use the simple biochemical data to generate the classification of the sample as positive or negative for liver fibrosis.

[0064] In an embodiment, the system is used to assess the presence of liver disease in an individual in a non-invasive manner. The system can provide a simple, inexpensive and accurate method for the diagnosis of advanced fibrosis, potentially avoiding the need for a liver biopsy in a significant percentage of patients. The system may be particularly useful in clinical settings where rapid and accurate assessment of liver fibrosis is required.

[0065] In an embodiment, the liver disease assessment method is implemented by instructions stored on a non-transitory computer-readable medium. The non-transitory computer-readable medium comprises hard disks, solid state drives, optical disks, flash memory devices, or any other type of storage device capable of storing data in a non-volatile manner. Preferably, the instructions stored on the non-transitory computer-readable medium may be executed by a processor, such as a central processing unit (CPU), a graphics processing unit (GPU), or any other suitable processing device.

[0066] In an embodiment, the instructions comprise steps for receiving the levels of a plurality of steroid biomarkers measured in a sample isolated from the individual. Preferably, the plurality of steroid biomarkers is as described in detail above.

[0067] Once the steroid biomarker levels are received, the instructions may include steps for feeding the received levels into the algorithm.

[0068] In an embodiment, the instructions stored on the non-transitory computer- readable medium are part of an application or software program installed on a computer or other electronic device. Preferably, the software application or program may provide an easy-to-use interface for entering steroid biomarker levels and clinical data and for displaying the classification generated by the algorithm. Preferably, the software application or program may also include functions for storing and managing data, for performing statistical analyses, and for generating reports or other outputs based on the results of the assessment of liver fibrosis.

[0069] In an embodiment, the liver disease assessment system comprises an input interface. This input interface can be configured to receive clinical data from the individual.

[0070] Preferably, the input interface is a physical interface, such as a keyboard or a touchscreen, or a software interface, such as a graphical user interface or a command line interface. The input interface can allow a user, for example a healthcare professional, to enter clinical data into the system in a practical and efficient way.

[0071] In an embodiment, the system comprises a processor. The processor can be configured to execute the algorithm that has been trained with a plurality of independent training samples. The processor may be a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor, or any other suitable processing device. The processor may be capable of executing the algorithm at high speed, thereby allowing rapid assessment of liver fibrosis.

[0072] In some aspects, the processor can be configured to receive steroid biomarker levels from the LC-MS / MS analyzer and clinical data from the input interface. The processor may then feed this data into the algorithm to generate a classification of the sample as positive or negative for liver fibrosis. The processor can also be configured to output the classification, for example by displaying it on a display device or by transmitting it to another device.

[0073] In an embodiment, the system may also include a memory device for storing the algorithm and the training samples. Preferably, the memory device can be a non-volatile memory device, such as a hard disk, a solid state drive or a flash memory device. The memory device can provide a secure and reliable storage solution for the algorithm and training samples, thus ensuring the integrity and availability of this data.

[0074] In an embodiment, a non-transitory computer-readable medium storing instructions for assessing liver fibrosis in an individual may be provided. When executed by a processor, these instructions may cause the processor to execute a method comprising receiving the levels of a plurality of steroid biomarkers measured in a sample isolated from the individual, feeding the received levels into an algorithm trained with a plurality of independent training samples, and generating, by the algorithm, a classification of the sample as positive or negative for liver disease. The non-transitory computer-readable medium can be a CD- ROM, DVD, flash drive, or any other suitable storage medium.

[0075] In an embodiment, the instructions stored on the non-transitory computer- readable medium may include further steps for receiving clinical data from the individual. The clinical data can be received by means of an input interface, such as a keyboard, touchscreen or graphic interface. The clinical data may include at least one of the following: age, body mass index (BMI), and presence or absence of type 2 diabetes mellitus. Simple biochemical data may include at least one of the following values: haemoglobin, platelets, ALT and albumin.

[0076] The method, system and non-transitory computer medium for the assessment of liver fibrosis by means of steroid biomarkers described herein have broad industrial applicability in the medical and healthcare field. The present invention provides a non-invasive and inexpensive approach for the diagnosis and monitoring of liver fibrosis, which can be implemented in clinical settings, diagnostic laboratories and research institutes. The use of advanced analytical techniques such as LC-MS / MS, combined with sophisticated algorithms, allows a more accurate and comprehensive assessment of liver health with respect to the traditional methods. While particularly useful in hepatology and gastroenterology, this approach can also find application in related fields such as endocrinology, internal medicine, and personalized medicine. Example

[0077] A method for serum dosing a steroid profile extended to 27 analytes has been developed (Tab. 1 ), using an advanced technology dosage (liquid chromatography coupled with mass spectrometry, LC-MS / MS), and this steroid panel was applied to a cohort of patients with MASLD, diagnosed by means of biopsy (313 subjects, 190 men and 123 women) and in healthy subjects (112 individuals, 55 men and 57 women).

[0078] Tab. 1 - Composition of the 27-analyte steroid panel studied in the reference cohort and in the validation cohort.

[0079] Analytes Abbreviations

[0080] Main compounds

[0081] Cortisone, ng / mL E

[0082] Cortisol, ng / mL F

[0083] 21 -Deoxy cortisol, pg / mL 21 -DF

[0084] Corticosterone, pg / mL CORT

[0085] 11 - Deoxy cortisol, pg / mL 11 -DF

[0086] Androstenedione, pg / mL A4

[0087] 11 -Deoxycorticosterone, pg / mL DOC

[0088] Testosterone, pg / mL T

[0089] Dehydroepiandrosterone, pg / mL DHEA

[0090] 17a-Hydroxyprogesterone, pg / mL 17aOHP

[0091] Epitestosterone, pg / mL EpiT

[0092] Dihydrotestosterone, pg / mL DHT

[0093] Progesterone, pg / mL P

[0094] Glucuronides (Glucuro-conjugates)

[0095] 5a-androstan-3a,17P-diol 3-glucuronide, pg / mL 5aa(3-Adiol 3G

[0096] 5(3-androstan-3a,17P-diol 3-glucuronide, pg / mL 5PaP-Adiol 3G

[0097] 5a-androstan-3a,17P-diol 17 -glucuronide, pg / mL 5aaP-Adiol 17G

[0098] 5P-androstan-3a,17P-diol 17 -glucuronide, pg / mL 5PaP-Adiol 17G

[0099] Etiocholanolone glucuronide, pg / mL EtioG

[0100] Androsterone glucuronide, pg / mL AG

[0101] Testosterone glucuronide, pg / mL TG

[0102] Sulphates (Sulphur-conjugates) Dehydroepiandrosterone sulphate, ng / mL DHEAS

[0103] Testosterone sulphate, pg / mL TS

[0104] Epitestosterone sulphate, pg / mL EpiTS

[0105] Epiandrosterone sulphate, ng / mL Epi AS

[0106] Etiocholanolone sulphate, ng / mL EtioS

[0107] Androsterone sulphate, ng / mL AS

[0108] From the results of the univariate analysis, numerous significant differences emerged between patients with MASLD and healthy subjects (Tab. 2) and, in the group of hepatopathic patients, among patients with absent / mild / moderate fibrosis (F0-F1 -F2) and with advanced fibrosis (F3- F4) (Tab. 3).

[0109] These differences were physio-pathologically consistent with what was expected and reproducible in the population even after stratification by sex (Tab. 3), although most of the changes observed concerned androgenic steroids.

[0110] Table 2 To better study these differences, all the variables collected and potentially associated with the outcome were included in a multivariate analysis model in logistic regression, which through stepwise backward selection, allowed the identification of clinical variables (age, BMI and diagnosis of type 2 diabetes mellitus), simple biochemical covariates and steroid analytes in logarithmic function, significantly associated with advanced liver fibrosis (classified as F3-F4 with liver biopsy).

[0111] Table 3

[0112] This analysis allowed to identify the analytes independently statistically significantly associated with advanced fibrosis (cortisol, cortisone, corticosterone, testosterone, androsterone glucuronide, androsterone sulphate, etiocholanolone glucuronide, etiocholanolone sulphate, 5a- androstan-3a , 17|3-diol 3-glucuronide and 5[3-androstan-3a,17[3-diol 3- glucuronide) (Tab. 4). Furthermore, the model demonstrated high accuracy for the diagnosis of advanced fibrosis (AUC 0.88, 95% Cl 0.84- 0.92). The model passed internal validation, through 10-fold cross validation, and was then validated on an external cohort, performing the same analyses on a large cohort of patients with histological diagnosis of MASLD, enrolled at the Division of Gastroenterology and Hepatology of the University of Palermo (145 patients, 81 male and 64 female). Also in this case, the model confirmed excellent stability and performance (AUC 0.82, 95% Cl 0.74-0.88), results comparable to what was observed in the reference cohort.

[0113] Table 4.

[0114] Coefficient P value IC 95%

[0115] Age > 50 years 0.6296096 0.128 - 0.1802721 1.439491

[0116] BMI 0.0865967 0.020 0.013348 0.1598453

[0117] Type 2 diabetes 1.182197 0.001 0.4538215 1.910572 mellitus

[0118] ALT (U / L) 0.0134803 0.002 0.0049077 0.022053

[0119] Haemoglobin (g / dL) - 0.3136743 0.043 - 0.6175535 - 0.009795

[0120] Albumin (g / dL) - 0.6654394 0.191 - 1.662233 0.3313543

[0121] Platelets (xlO9 / L) - 0.0101898 0.001 - 0.0163423 - 0.0040373

[0122] LOG F 2.558834 0.069 - 0.1951208 5.312789

[0123] LOG E - 2.572535 0.005 - 4.347488 - 0.7975817

[0124] LOG CORT - 1.34031 0.063 - 2.753416 0.072797

[0125] LOG T 0.7478668 0.120 - 0.1938627 1.689596

[0126] LOG AG 3.437295 0.001 1.402934 5.471657

[0127] LOG AS -2.36868 < 0.001 -3.690281 -1.04708

[0128] LOG EtioG - 1.745439 0.050 - 3.487339 - 0.0035391

[0129] LOG EtioS 1.539441 0.011 0.3584863 2.720395

[0130] LOG 5aabl7G - 1.171636 0.071 - 2.442052 0.0987799

[0131] LOG 5aab3G - 1.314441 0.136 - 3.041173 0.4122913

[0132] Constant 8.9787 0.030 0.8561302 17.10127 In light of the results achieved, an algorithm (which we called Fibrosis and Steroid [Faster] - model) was derived that could be applied to each patient suffering from MASLD, in order to diagnose advanced fibrosis in a simple, economic and non-invasive manner (combining some patient data with the dosages performed on a simple blood sample), thus allowing to avoid liver biopsy in 80% of patients. The algorithm behind FaSter-model is as follows:

[0133] Y = (Age > 50 years * 0.6296096) + (BMI [Kg / m2] * 0.0865967) + (DMT2 * 1.182197) + (ALT [U / L] * 0.0134803) - (Haemoglobin [g / dL] * 0.3136743) - (Albumin [g / dL] * 0.66543949) - (Platelets [x109 / L] * 0.0101898) + (LOG F * 2.558834) - (LOG E * 2.572535) - (LOG CORT * 1.34031 ) + (LOG T * (0.7478668) + (LOG AG * 3.437295) - (LOG AS * 2.36868) - (LOG EtioG * 1.745439) + (LOG EtioS * 1.539441 ) - (LOG 5aab17G * 1.171636) - (LOG 5aab3G * 1 .314441 ) + 8.9787

[0134] (diagnosis of T2DM= 1 and absence of T2DM=0)

[0135] The Y value, derived from the developed algorithm, allows to diagnose:

[0136] • advanced liver fibrosis (Y > 2.72; Sensitivity 16.5%, Specificity

[0137] 100%)

[0138] • the absence of advanced liver fibrosis (Y < -3.39; Sensitivity 100%, Specificity 17.5%)

[0139] • diagnostic doubt (-3.39 < Y < 2.72), for which a liver biopsy could be indicated, in the opinion of the clinician.

Claims

AMENDED CLAIMS received by the International Bureau on 26 March 2026 (26.03.2026)1. An in vitro method for assessing the presence of liver fibrosis in an individual comprising the steps of: a) measuring in a sample isolated from the individual the concentration values of a plurality of biomarkers selected from the group consisting of: cortisone, cortisol, 21 -deoxycortisol, corticosterone, 11 -deoxycortisol, androstenedione, 11 -deoxycorticosterone, testosterone, dehydroepiandrosterone, 17a-hydroxyprogesterone, epitestosterone, dihydrotestosterone, progesterone, 5a-androstan-3a 17|3-diol 3-glucuronide, 5[3-androstan-3a 17|3-diol 3- glucuronide, 5a-androstan-3a 17|3-diol 17-glucuronide, 5[3-androstan-3a 17|3-diol 17- glucuronide, etiocholanolone glucuronide, androsterone glucuronide, testosterone glucuronide, dehydroepiandrosterone sulphate, testosterone sulphate, epitestosterone sulphate, epiandrosterone sulphate, etiocholanolone sulphate and androsterone sulphate; b) analyzing the values obtained in step a) by means of an algorithm trained to classify said sample as positive or negative for liver fibrosis, wherein the algorithm is a logistic regression model in which each concentration value obtained in step a) is associated with a coefficient obtained by training the algorithm.

2. The method according to claim 1 , wherein the biomarkers are at least six biomarkers selected from the group consisting of following biomarkers: cortisol, cortisone, corticosterone, testosterone, androsterone glucuronide, androsterone sulphate, etiocholanolone glucuronide, etiocholanolone sulphate, 5a-androstan-3a,17[3-diol 3- glucuronide and 5[3-androstan-3a,17[3-diol 3-glucuronide.

3. The method according to claim 1 , wherein the biomarkers are cortisol, cortisone, corticosterone, testosterone, androsterone glucuronide, androsterone sulphate, etiocholanolone glucuronide, etiocholanolone sulphate, 5a-androstan-3a,17[3-diol 3- glucuronide and 5[3-androstan-3a,17[3-diol 3-glucuronide.

4. The method according to claim 1 , wherein the biomarkers are at least six biomarkers selected from the group consisting of following biomarkers: cortisol, cortisone, 17a- hydroxyprogesterone, testosterone, androsterone glucuronide, androsterone sulphate, etiocholanolone glucuronide and etiocholanolone sulphate.

5. The method according to claim 1 or 4, wherein the biomarkers are cortisol, cortisone, 17a- hydroxyprogesterone, testosterone, androsterone glucuronide, androsterone sulphate, etiocholanolone glucuronide and etiocholanolone sulphate.

6. The method according to any one of claims 1 to 5, wherein the training of the algorithm is obtained by processing some values of the biomarkers obtained from a plurality of independent training samples from said sample, said training samples being isolated from individuals having liver fibrosis and individuals not having liver fibrosis.

7. The method according to any one of claims 1 to 6, wherein the sample is selected from blood, plasma or serum.

8. The method according to any one of claims 1 to 7, wherein the individual suffers from hepatic steatosis.

9. The method according to claim 8, wherein the individual has hepatic steatosis associated with metabolic dysfunction (MASLD) or steatohepatitis associated with metabolic dysfunction (MASH).

10. The method according to any one of claims 1 to 9, wherein the method comprises a step of evaluating clinical data of the individual selected from: age, Body Mass Index (BMI) and the presence or the absence of type 2 diabetes mellitus and a step of evaluating simple biochemical data selected from: haemoglobin, platelets, ALT and albumin.

11. The method according to claim 10, wherein each clinical data is associated with a coefficient obtained by the training of the algorithm, wherein the training of the algorithm is obtained by processing clinical data obtained from a plurality of independent training samples from said sample, said training samples being isolated from individuals with liver fibrosis and from individuals without liver fibrosis.

12. The method according to any one of claims 1 to 11 , wherein the logistic regression model comprises a constant, preferably comprised between 7 and 10.

13. The method according to any one of claims 1 to 12, wherein the cortisol is expressed as the base 10 logarithm of the nanograms per milliliter (ng / mL) of the cortisol detected in the individual and wherein the cortisol coefficient is between 2.3 and 2.8, preferably between 2.45 and 2.65.

14. The method according to any one of claims 1 to 13, wherein the cortisone is expressed as the base 10 logarithm of the nanograms per milliliter (ng / mL) of the cortisone detected in the individual and wherein the cortisone coefficient is comprised between -2.3 and -2.8, preferably between -2.45 and -2.65.

15. The method according to any one of claims 1 to 14, wherein the corticosterone is expressed as the base 10 logarithm of the picograms per milliliter (pg / mL) of the corticosterone detected in the individual, and wherein the corticosterone coefficient is comprised between -1 .0 and -1 .7, preferably between -1 .2 and -1 .5.

16. The method according to any one of claims 1 to 15, wherein the testosterone is expressed as the base 10 logarithm of the picograms per milliliter (pg / mL) of the testosterone detected in the individual and wherein the testosterone coefficient is comprised between 0.4 and 1.1 , preferably between 0.6 and 0.9.

17. The method according to any one of claims 1 to 16, wherein the androsterone glucuronide is expressed as the base 10 logarithm of the pg / mL of the androsterone glucuronide detected in the individual, and wherein the androsterone glucuronide coefficient is comprised between 2.9 and 4, preferably between 3.2 and 3.7.

18. The method according to any one of claims 1 to 17, wherein the androsterone sulphate is expressed as the base 10 logarithm of the ng / mL of the androsterone sulphate detected in the individual and wherein the coefficient of androsterone sulphate is comprised between -1 .8 and -2.9, preferably between -2 and -2.7.

19. The method according to any one of claims 1 to 18, wherein the etiocholanolone glucuronide is expressed as the base 10 logarithm of the pg / mL of the etiocholanolone glucuronide detected in the individual, and wherein the coefficient of the etiocholanolone glucuronide is comprised between -1 .2 and -2.3, preferably between -1 .4 and -2.1 .

20. The method according to any one of claims 1 to 19, wherein the etiocholanolone sulphate is expressed as the base 10 logarithm of the picograms per milliliter (pg / mL) of the etiocholanolone sulphate detected in the individual, and wherein the coefficient of the etiocholanolone sulphate is comprised between 1.3 and 1.9, preferably between 1.4 and 1.6.

21. The method according to any one of claims 1 to 20, wherein the 5a-androstan-3a,17[3- diol 3-glucuronide is expressed as the base 10 logarithm of the pg / mL of the 5a-androstan- 3a, 17|3-diol 3-glucuronide detected in the individual, and wherein the coefficient of the 5a- androstan-3a,17[3-diol 3-glucuronide is comprised between -0.7 and -1.6, preferably between -0.9 and -1 .4.

22. The method according to any one of claims 1 to 21 , wherein the 5a-androstan-3a,17[3- diol 17-glucuronide is expressed as the base 10 logarithm of the pg / mL of the 5a-androstan- 3a, 17|3-diol 17-glucuronide detected in the individual, and wherein the coefficient of the 5a- androstan-3a,17[3-diol 17-glucuronide is comprised between -0.8 and -1.83, preferably between -1.1 and -1 .53.

23. A system for assessing a liver disease in an individual, characterized by the fact that the system comprises: a liquid chromatography and tandem mass spectrometry (LC-MS / MS) analyzer configured to measure the levels of a plurality of steroid biomarkers in a sample isolated from an individual, wherein the plurality of steroid biomarkers comprises the following biomarkers: cortisol, cortisone, corticosterone, testosterone, androsterone glucuronide, androsterone sulphate, etiocholanolone glucuronide, etiocholanolone sulphate, 5a-androstan-3a,17[3-diol 3-glucuronide and 5[3-androstan-3a,17[3-diol 3-glucuronide; and a processor capable of executing an algorithm trained with a plurality of independent training samples, wherein said algorithm is configured to generate a classification of said sample as positive or negative for liver fibrosis based on the measured levels of the plurality of steroid biomarkers.

24. The system according to claim 23, wherein the algorithm is a logistic regression model according to claims 1 -22.

25. A non-transitory computer-readable medium containing instructions suitable for inducing the system of claim 23 or 24 to perform the method according to any one of claims 1 -22.

Images