Clinical and molecular prognostic markers in liver transplantation

Abstract

The present invention relates to a method for predicting the outcome of liver transplantation for the treatment of hepatocellular carcinoma (HCC), or a method for arranging HCC patients to receive liver transplantation, comprising the following steps: determining the expression levels of indicator genes (including skin pontin, clusterin, calpain small subunit 1, F-box and WD repeat protein 7 and SproutyRTK signaling antagonist 2) selected from liver samples of patients suffering from HCC, and comparing the expression levels of the indicator genes with the expression levels of internal reference genes, and combining this with the variable total tumor volume through a linear support vector machine algorithm to predict a good prognosis for the liver transplant recipient.

Description

[0001] This application claims priority to European Patent Application EP19201218.5, dated October 2, 2019, which is incorporated herein by reference. The present invention relates to a method for assessing the prognosis or determining the likelihood of a positive outcome in patients pre-selected for or considered for liver transplantation during the treatment of hepatocellular carcinoma. The integrated application of clinical and molecular markers allows for better selection of potential liver transplant candidates compared to current patient selection criteria. Its application will contribute to improving liver transplant selection for patients with hepatocellular carcinoma.

[0002] manual

[0003] Hepatocellular carcinoma (HCC) is a highly prevalent disease that significantly impacts mortality and quality of life. Understanding its multifactorial etiology and complex pathogenesis is crucial to providing patient stratification information for current treatments or designing personalized medicines optimized for individual patient genomes. Liver transplantation (LT) is the optimal treatment for HCC in cirrhotic patients, but organ availability is limited due to the disease's high likelihood of adverse outcomes. Furthermore, it is believed that expanding current standards will increase the demand for transplants, leading to longer waiting times on pre-transplant lists, higher dropout rates, and worsening intended treatment outcomes. Even when organs are available, the benefits of LT must be balanced against the risks to the donor.

[0004] Therefore, optimizing the selection of HCC patients is imperative. The strictness of current clinical morphological models, namely the Milan criteria (MC), may exclude excellent candidates, such as patients with advanced yet "benign" tumors discovered at a late stage of the disease. Conversely, MC may not exclude patients with early-stage HCC characterized by more aggressive tumor behavior. Identifying optimal liver transplant candidates for patients with HCC in the context of cirrhosis could increase the number of LT candidates, thereby impacting the demand for donors.

[0005] Molecular biomarkers can provide information about the biological behavior of tumors. However, there are currently no universally accepted prognostic biomarkers for the proper selection of transplant patients, as few studies have used HCC molecular biomarkers in patient care. The inventors propose that it is important to focus on genes or genetic characteristics that influence prognosis, as tumor morphology-based criteria can only partially differentiate patients into those with good and poor prognoses. Genes expressed in HCC patients with good prognoses (i.e., genes upregulated leading to good outcomes) can supplement currently accepted criteria for selecting subgroups of patients with a high probability of benefiting from transplantation (LT). Of course, negative predictive genes also play an important role, as they can identify patients with a high likelihood of disease recurrence.

[0006] Dermatopontin (DPT)

[0007] DPT, also known as TRAMP (tyrosine-rich acidic matrix protein) (gene ID GC01M168664), is an extracellular matrix protein that may play a role in cell-matrix interactions and matrix assembly. This protein is present in various tissues and is thought to be expressed in mesenchymal cells (fibroblasts and myofibroblasts) and macrophages. This molecule is essential for extracellular matrix assembly, cell adhesion, and wound healing. It also accelerates collagen fibrillation and alters the behavior of TGF-β through interaction with core proteoglycans in the extracellular matrix microenvironment in vivo. DPT inhibits the formation of the core proteoglycan TGF-β1 complex and may increase cellular responses to TGF-β, enhancing its biological activity. Furthermore, it has been identified as a downstream target of the vitamin D receptor. The vitamin D receptor, in turn, mediates downstream signaling of 1,25-dihydroxyvitamin D3, thereby exerting an antiproliferative effect against HCC. DPT is associated with cell adhesion and tumor invasiveness. Strong expression of DPT is associated with the inhibition of metastasis in oral cancer and giant cell tumor of bone. Downregulation of DPT is associated with carcinogenesis and progression of HCC through possible interaction with TGF-β1 and other potential mechanisms. DPT expression levels in HCC tissues are significantly lower than in healthy livers.

[0008] Calonase small subunit 1 (CAPNS1)

[0009] Caloplasm is a family of proteases that participate in cell migration and invasion by altering the structure of cell adhesion molecules and cytoskeleton components or by interfering with intracellular signaling pathways. Its regulatory subunits, CAPNS1 and CAPN4, may play important roles in caloplasm activity. Knockdown of CAPNS1 in malignant endothelial cells may reduce the proliferative capacity of malignant endothelial cells, while upregulation of CAPNS1 is associated with increased tumor size, number, and alpha-fetoprotein (AFP) levels in animals after HCC resection.

[0010] Cluster protein (CLU)

[0011] The protein encoded by the CLU gene chaperone (CLU, gene ID GC08M027596) is cytoplasmic and secretory. CLU complexed with EIF3I activates the Akt pathway, thereby promoting the metastasis of HCC cells.

[0012] Contains F-box and WD repeat protein 7 (FBXW7, gene ID GC04M152321).

[0013] The gene “F-box and WD repeat domain protein 7 (FBXW7),” also known as Sel10, hCDC4, or hAgo, encodes a member of the F-box protein family, which serves as a substrate recognition component for the SCFE3 ubiquitin ligase. FBXW7 is a key tumor suppressor and one of the most commonly dysregulated ubiquitin-proteasome system proteins in human cancers. FBXW7 controls the degradation of proteasome-mediated oncoproteins such as cyclin E, c-Myc, Mcl-1, mTOR, Jun, Notch, and AURKA. Mutations in this gene have been detected in ovarian and breast cancer cell lines, suggesting a potential role in human cancer pathogenesis. In vitro studies have shown that FBXW7 can serve as a prognostic marker for hepatocellular carcinoma (HCC), and lower FBXW7 expression levels are associated with lower survival rates in HCC patients. Therefore, FBXW7 plays a crucial role in HCC progression, as it inhibits HCC cell migration and invasion through the Notch1 signaling pathway.

[0014] Sprouty RTK signaling antagonist 2 (SPRY2, gene ID GC13M080335)

[0015] The Drosophila Spry (dSPRY) gene family includes four homologs, SPRY1 through SPRY4, which are believed to be involved in the negative feedback loop of the RAF / MEK / ERK pathway associated with HCC carcinogenesis. In particular, SPRY2 antagonizes growth factor-mediated cell proliferation, migration, and differentiation by regulating receptor tyrosine kinase (RTK) signaling and inhibiting the RAF / MEK / ERK pathway. This protein is a crucial regulator of important pathways related to cancer development, such as angiogenesis, cell growth, invasion, migration, and cytokinesis.

[0016] Based on the aforementioned latest technology, the object of this invention is to provide molecular biomarkers that facilitate the accurate selection of optimal candidates for tumor LT from all HHC patients. These molecular biomarkers should be able to predict tumor behavior and invasiveness. This object is achieved through the claims of this specification as a whole and the further advantageous embodiments provided herein. The inventors have refined the clinical criteria for selecting HCC patients with cirrhosis for LT and evaluated the role of selected biomarkers in the population. The inventors have validated the role of current clinical biomarkers to specifically identify optimal candidates for LT in subgroups of patients outside the current clinical gold standard of the Milan criteria, thereby integrating clinical and molecular characteristics to accurately address the biological problems of tumors.

[0017] A set of potential candidate genes was identified using clinical data from public repositories, systematic reviews on molecular prognostic biomarkers for HCC, and internal clinical data available to the inventors. The inventors identified single-gene alterations with prognostic value and combined them into predictive multivariate traits. The systematic reviews were able to identify genes associated with HCC progression, which are believed to provide prognostic information after LT, thus aiding in patient selection for LT.

[0018] The inventors here demonstrate that DPT, CLU, CAPNS1, FBXW7, and SPRY2 exhibit differential expression in patients with or without HHC recurrence after LT. Furthermore, DPT and CLU, alone or in combination, effectively differentiate a subgroup of patients with non-recurrent HCC after LT, thus indicating a favorable prognosis.

[0019] The term "gene expression" or "expression," or the term "gene product," can refer to the process and products of either the production of nucleic acids (RNA) or the production of peptides or polypeptides (also referred to as transcription and translation, respectively), or any intermediate process that regulates the processing of genetic information to produce a polypeptide product. The term "gene expression" can also be used for the transcription and processing of RNA gene products, such as regulatory RNA or structural RNA (e.g., ribosomal RNA). If the expressed polynucleotide originates from genomic DNA, the expression may include splicing of mRNA in eukaryotic cells. Expression can be measured at the transcriptional and translational levels, also known as mRNA and / or protein products.

[0020] In the context of this invention, the term "good prognosis" refers to the absence of HHC disease recurrence within five years after LT. In embodiments, good prognosis is measured by its direct correlation with patient survival, and therefore overall survival or disease-free survival is largely equivalent.

[0021] In the context of this invention, the terms "Support Vector Machine," "SVM," "Linear Kernel SVM," or "SVM algorithm" refer to a supervised machine learning model capable of classifying data and / or performing regression analysis. It is sometimes referred to as a Support Vector Network and, in the context of this invention, is used as a form of binary linear classification algorithm. In the context of this invention, the algorithm uses a training step where patient data samples are associated with a set of variables to build a model, including but not limited to gene expression levels or tumor volume measurements, which assigns samples to one or another category, such as survival rate or five-year survival rate.

[0022] On one hand, the present invention relates to a method for predicting liver transplant outcomes to treat HCC. An alternative aspect of this invention relates to a method for treating HCC via liver transplantation. In another alternative aspect, the present invention relates to a method for grouping HCC patients into different groups who are more or less likely to benefit from receiving a liver transplant; in other words, these patients will exhibit longer overall survival after liver transplantation without disease recurrence.

[0023] The method according to this aspect of the invention includes the step of determining the expression level (particularly mRNA level) of a genetic biomarker or indicator gene (the expression level of the indicator gene is also referred to herein as the indicator gene expression level) in a liver sample obtained from a patient suffering from HCC, wherein the indicator gene is selected from a list including:

[0024] a. Dermatoponins;

[0025] b. Cluster protein;

[0026] c. Calpain small subunit 1;

[0027] d. Contains F-box and WD repeat protein 7;

[0028] e.Sprouty RTK signaling antagonist 2.

[0029] In some implementations, the indicator gene is CLU or DPT.

[0030] In an alternative aspect to the first aspect of the invention, the method for predicting liver transplant outcomes to treat HCC comprises:

[0031] - In the determination step, the expression level of each indicator gene, including CLU and DPT, is determined in liver samples obtained from patients suffering from HCC;

[0032] - In the classification step, a good prognosis is designated as the outcome of liver transplantation based on the expression level of indicator genes.

[0033] In some implementations, this set of indicator genes also includes at least one of the following:

[0034] -Calcase small subunit 1;

[0035] - Contains F-box and WD repeat protein 7;

[0036] -Sprouty RTK signaling antagonist 2.

[0037] In some implementations, overexpression of any indicator gene indicates a good prognosis.

[0038] In some implementations, overexpression of DPT and / or CLU indicates a favorable prognosis.

[0039] In some implementations, the indicator gene is overexpressed relative to a threshold. In alternative implementations, the indicator gene expression level is compared to a control sample selected from representative patients in each disease outcome subset.

[0040] In some embodiments, the expression level of the indicator gene is determined using a quantitative polymerase chain reaction (PCR) sensitive to the mRNA levels encoding the indicator gene present in the sample. Table 6 shows specific primers for amplifying the target indicator gene region specified in this invention, which can be specifically used to determine the expression level of the indicator gene according to the method provided in this invention. Global RNA sequencing is an alternative method that can produce gene expression levels used according to this invention.

[0041] The expression levels of indicator genes can be compared with those of internal reference genes, particularly with the expression levels of housekeeping genes. In a particular embodiment, this is compared with the expression level of ribosomal protein L13A (RPL13A) (gene ID 23521). Other possibilities include other housekeeping genes, such as GADPH and / or TBP, or combinations thereof.

[0042] In some implementations, indicator gene expression levels are determined by PCR. The indicator gene expression value relative to a threshold is determined as the difference between the threshold cycle number of the indicator gene and the threshold cycle number of the internal reference gene. The threshold cycle number refers to the number of PCR cycles required to detect the products [the indicator gene and the internal reference gene].

[0043] In some implementations, the following calculations are used to obtain values ​​reflecting the expression level of indicator genes:

[0044] Target quantity = ACt

[0045] ACt(R) = Ct(DPT in R) - Ct(reference gene in R)

[0046] ACt(nonR) = Ct(DPT in nonR) - Ct(reference gene in nonR)

[0047] Where R represents samples with disease recurrence, and nonR represents samples without disease recurrence.

[0048] In some implementations, the indicator gene is DPT, and the threshold is gene expression difference or ACt, which is higher than 7.

[0049] In some implementations, the indicator gene is CLU, and the threshold is gene expression difference or ACt, which is higher than -0.54.

[0050] In some implementations, the difference in the threshold cycle number of the indicator gene is used, and genes above this difference are said to be overexpressed. The difference is: DPT expression level ACt value greater than 7, CLU expression level ACt value greater than -0.54.

[0051] The thresholds provided here are examples of ACt values, which can be used to classify gene expression levels according to the present invention. They reflect gene expression relative to housekeeping genes. Positive gene expression levels indicate that the gene is expressed more than the housekeeping reference gene, while negative expression values ​​indicate that the gene is expressed less than the housekeeping reference gene. Overexpression of indicator genes is defined as expression above the ACt threshold, which distinguishes patients with poor prognosis from those with good prognosis.

[0052] In some implementations, DPT overexpression and a total tumor volume ≤115 cm³ are preferred. 3 The test results indicate a good prognosis.

[0053] In some implementations, CLU overexpression and a total tumor volume ≤115 cm³ are considered optimal. 3 The test results indicate a good prognosis.

[0054] In some embodiments, multiple patient factors or variables are used to predict patient outcomes, such as the expression levels of CLU, DPT, and tumor volume. In a specific embodiment of this aspect of the invention, the expression levels of the predictive indicator genes CLU and DPT exceeding a specified threshold indicate a favorable prognosis. In a related embodiment, indicator gene expression and HHC tumor volume ≤115 cm⁻¹ are considered favorable. 3 Additional non-genetic variables indicate a good prognosis.

[0055] In some implementations, the expression levels of one or more indicator genes and / or tumor volume measurements in a patient sample are incorporated into the algorithm to provide values ​​reflecting the likelihood of disease recurrence, particularly where the algorithm is a support vector machine algorithm, and more particularly where the algorithm is a linear kernel support vector machine algorithm.

[0056] In some implementations, indicator gene expression levels and tumor size variables are used in predictive algorithms, particularly machine learning algorithms, and more particularly linear kernel support vector machine (SVM) learning algorithms, which categorize patients, especially in the next five years after LT, into subsets that are likely or unlikely to have a good prognosis.

[0057] The data provided in the examples categorized CLU overexpression into Ct ranges from 27.36 to 40.5 and DPT ranges from 33.49 to 35.69 in the cohort. -0.54 was represented as the change in Ct compared to the control group (also known as delta or ACt) and thus determined as a useful threshold for CLU overexpression. Similarly, an ACt threshold of 7 was determined as a useful threshold for DPT expression in patient samples. Tumor volumes less than or equal to 115 cm⁻¹ were also found to be suitable for this cohort. 3 This also indicates a good prognosis. Tumor volume refers to the total volume of all tumors within the patient's liver; useful methods for determining tumor volume can be selected from computed tomography (CT) and magnetic resonance imaging (MRI). Thresholds, such as those specified in this aspect of the invention, can be used specifically for binarizing patient values ​​to aid in the classification of patient outcomes. The data in Example 5 provide useful thresholds for CLU, DPT, and tumor volume. Samples with expression levels or volumes below this value are assigned a score of 0, and samples with measurements above this threshold are assigned a score of 1. The examples demonstrate incorporating binarized multivariate data from the patient cohort into an SVM algorithm to create a classification system where the resulting scores represent patient outcomes after LT.

[0058] In some implementations, the determination step includes determining the expression of DPT and the expression of genes selected from CLU, CAPNS1, FBXW7, and SPRY2, particularly the expression of DPT and CLU.

[0059] In some implementations, overexpression of either CLU or DPT alone indicates a favorable prognosis. In some implementations, the predictive classification based on indicator gene expression levels and / or TTV, or both, provided by this invention is combined with additional prognostic predictive factors assessed under, for example, the Milan criteria. Data from the examples demonstrate that CLU expression levels can accurately predict disease outcomes in a subset of HHC patients classified outside the Milan criteria.

[0060] The Milan criteria were introduced by Mazzaferro in 1996 (Mazzaferro et al, N Engl J Med. 1996 Mar 14; 334(11): 693-9). The criteria impose the following restrictions on transplantation in adults with HCC: (1) the diameter of a single tumor is less than 5 cm; (2) the number of tumor lesions does not exceed three, and the size of each lesion does not exceed 3 cm; (3) there is no vascular invasion; and (4) there is no extrahepatic involvement.

[0061] The present invention also relates to a system for detecting high expression of liver transplant biomarkers. The system includes a device for determining the expression of DPT and the expression of genes selected from CLU, CAPNS1, FBXW7, and SPRY2, particularly the expression of DPT and CLU.

[0062] On the other hand, the present invention includes the use of primers for amplifying and detecting the expression of DPT and CLU, and optionally other biomarkers selected from CLU, CAPNS1, FBXW7, and SPRY2, in a kit for analyzing biomarkers to predict the outcome of liver transplantation for the treatment of HCC. Furthermore, the present invention includes a method for treating a patient previously diagnosed with HHC via liver transplantation, wherein the method described according to any of the foregoing aspects and embodiments has classified the patient as potentially having a good prognosis. Attached Figure Description

[0063] Figure 1 The results of RT-qPCR for differential expression of genes selected in the validation set based on relapse rate are shown. Relapse (R: left) vs. no relapse (noR: right). Wilcoxon (Mann-Whitney) test (confidence level = 0.95).

[0064] Figure 2 The receiver operating characteristic (ROC) analysis for DPT expression and tumor recurrence is shown. The area under the curve (AUC) was 0.77. A cutoff value of 7 was selected for ACt, corresponding to 84% sensitivity and 63% specificity.

[0065] Figure 3 The disease-free survival rate obtained from DPT expression is shown.

[0066] Figure 4 shows the survival rates obtained from DPT expression in patients within and outside the Milan criteria (MC).

[0067] Figure 5 shows the situation where TTV > 115cm 3 Disease-free survival rates derived from DPT expression in patients (A), poorly differentiated patients (B), or patients with microvascular invasion (C).

[0068] Figure 6 The survival rate obtained from DPT expression in the HCV subgroup of patients is shown.

[0069] Figure 7 The disease-free survival rate obtained from CLU expression is shown.

[0070] Figure 8 shows disease-free survival rates based on CLU expression in the “poor prognosis” subgroup: A. Patients outside the MC; B. TTV > 115cm 3 Patients with: A. Microvascular invasion; B. Poorly differentiated tumors; C. Microvascular invasion; D. Poorly differentiated tumors.

[0071] Figure 9 The survival rate is shown based on CLU expression in hepatitis C virus (HCV) patients.

[0072] Figure 10 The disease-free survival rate is shown as obtained from the DPT / CLU composite score.

[0073] Figure 11 shows a patient outside the MC (A) with a TTV > 115cm. 3 DFS was obtained using a composite genetic score in patients (B), patients with poorly differentiated tumors (C), and patients with microvascular invasion (D).

[0074] Figure 12 Showing for TTV < 115cm 3 DFS of patients with composite genetic scores.

[0075] Figure 13 Disease-free survival is shown based on the expression of DPT and / or CLU. Compared with CLU, DPT expression showed a better ability to predict long-term survival.

[0076] Figure 14 Confusion matrix of the prediction algorithm.

[0077] Figure 15 The Kaplan-Meier curves show the disease-free survival rates of the total population in the cohort based on an algorithm using CLU, DPT, and TTV values.

[0078] Figure 16 The Kaplan-Meier curves show disease-free survival rates for patients outside the Milan criteria based on an algorithm using CLU, DPT, and TTV values.

[0079] Figure 17 The Kaplan-Meier curves show disease-free survival rates for patients within the Milan criteria based on an algorithm using CLU, DPT, and TTV values. Example

[0080] Example 1: Study of molecular prognostic biomarkers

[0081] The inventors analyzed the performance of several current selection criteria based on morphological characteristics, highlighting their potential to exclude good candidates and incorrectly include poor candidates. The inventors tested previously identified genes as presumptive biomarkers for HCC prognosis after transplantation. The DPT and CLU genes effectively distinguished, isolated, or both of these genes in a subgroup of patients with extremely low recurrence rates after HCC liver transplantation.

[0082] Example 2: Prognostic molecular markers for hepatocellular carcinoma

[0083] Because important information can be obtained from the patient cohort undergoing hepatectomy (LR), biomarker studies in LR were included in the analysis. Unlike single genes, overlap between traits is not common in the literature, and gene traits are often not reproducible. For this reason, the inventors focused specifically on single-gene biomarkers rather than gene traits. Furthermore, since many biomarkers have not been further tested and validated outside their original research settings, their use in this current work at least represents the result of external validation efforts.

[0084] The inventors retrieved the following data: data types (mRNA, miRNA, and protein), prognostic information, specific genes involved, genes with good or bad prognostic outcomes, alteration types (overexpression, downregulation, high / low methylation, mutation), patient samples and statistical data, and the authors' observations.

[0085] Example 3: Experimental Set

[0086] The first set was implemented to better prune biomarkers associated with better prognosis identified throughout the literature search and to further reduce the number of putative biomarkers associated with early relapse. From the initially proposed 20 patients (see Example 6, Sample Collection), 9 more patients were added to the test set due to difficulties in extracting RNA from the first batch of samples from elderly patients. Samples from 3 patients were not used due to insufficient sample size. Therefore, the test set consisted of 26 patients (6 patients without relapse outside the MC; 7 patients with relapse outside the MC; 7 patients without relapse within the MC; and 6 patients with relapse within the MC). After RNA extraction, cDNA was obtained and RT-qPCR was performed. Finally, differential expression of the assessed genes was correlated with clinical data. Table 1 shows the test set results based on relapse status within or outside the MC.

[0087]

[0088] Table 1: Gene expression results are based on the test set and its relationship with relapse. p-values ​​are derived from the Mann-Whitney test.

[0089] In patients within the migratory macerator (MC), the expression of Clusterin (CLU, p = 0.09), CAPNS1 (p = 0.05), and FBXW7 (p = 0.04) differed significantly according to relapse status. In patients outside the MC, the expression of Clusterin (p = 0.02), Dermatopontin (p = 0.06), and SLC16A4 (p = 0.08) also differed significantly according to relapse status. Clusterin was the best biomarker for distinguishing overall relapse status (p = 0.01). A subset of selected biomarkers CAPNS1, DPT, CLU, and FBXW7 that met the selection criteria for the trial study were then applied to the validation set. Furthermore, SPRY2 was included in the validation set for further downstream analysis because its p-value was critically significant in the "within the Milan criteria" group of the trial set (p = 0.13), although it was not significant in the other two groups. Another gene, MUC15, which also had a critical significance p-value (p = 0.14) in the "All Standards" group, was excluded from the genomes to be tested in the validation set. This exclusion was due to its poor performance in RT-qPCR (excessively high Ct and frequent sample failures), which may be related to its extremely low expression levels.

[0090] Example 4: Validation Set

[0091] A complete population of 301 patients who underwent total liver transplantation (LT) for HCC between September 1992 and February 2014 was considered. Based on inclusion and exclusion criteria (see Example 6, Study Population), a total of 275 patients with histologically confirmed HCC were identified in this analysis. Of the 44 excluded patients, 32 had perioperative mortality and 12 had residual and / or extrahepatic disease. Patients previously included in the trial set (26) and those within 5 years post-LT (38) were also excluded. The initial study population consisted of 167 patients. 33 patients (19.7%) were ineligible for sample analysis due to extensive tumor necrosis (14 / 167; 8.3%), low-quality RNA extraction (12 / 167; 7.2%), or unavailable formalin-fixed paraffin-embedded tissue (7 / 167; 4.2%). Therefore, 20 patients within 5 years post-LT were added, bringing the total number of patients in the validation set to 154. The total number of patients who underwent molecular marker analysis was 180. Table 2 summarizes the general characteristics of the trial set (n=26), validation set (n=154), and total population (n=180).

[0092] As expected, the trial set exhibited a more aggressive tumor profile compared to the validation set, as patients were evenly distributed within and outside the MC in this first set. Patients in the trial set showed more frequent vascular invasion, larger tumors, higher TTV, and a lower proportion of patients within the MC. In the overall population, the median donor age was 38 years, consistent with a high proportion of FAP donors (patients with familial amyloid polyneuropathy receiving living donor transplants whose livers are available to treat other patients). VHC predominated among patients with infection, accounting for 46%. An important observation was the short wait time before LT—median 1 month, corresponding to a mean wait time of 2.2 (0–18) months. The median tumor size was 1, corresponding to a mean of 2.17 (1–11), and the largest tumor size was 2.75 cm (mean 3.25 cm, ranging from 0.3 cm to 20 cm). Only 120 (66.7%) patients were within the MC, and 154 (85.6%) had a TTV <115 cm. 3 Inside.

[0093]

[0094] Table 2: Comparison of demographic and clinical / pathological characteristics of the trial set, validation set, and total population. Comparisons are made only between the trial set and the validation set. IQR: interquartile range; AFP: alpha-fetoprotein; MELD: end-stage liver disease model.

[0095] Expression levels of CAPNS1, DPT, CLU, FBXW7, SLC16A4, and SPRY2 were analyzed in the validation set. In a large sample, SLC16A4 expression was too low to be identified, and further analysis of this specific biomarker was abandoned. Figure 1 The differential expression of genes selected based on recurrence rate is shown using the Wilcoxon (Mann-Whitney) test.

[0096] Calonase small subunit 1 (CAPNS1)

[0097] CAPNS1 was included in the validation set because it showed marginally different expression in patients with recurrence within the Milan criteria in the trial set (p = 0.05). Therefore, CAPNS1 expression levels in tumor tissues differed significantly based on recurrence in the validation set (OR 1.448, CI 1.140–1.840, p = 0.002). However, this did not translate into corresponding significance in hazard ratios, neither for DFS (HR 1.073, CI 0.978–1.178, p = 0.136) nor for OS (HR 1.014, CI 0.913–1.125, p = 0.796). ROC curves were used to determine the cutoff value for recurrence rate. The optimal cutoff value appeared to be a deltaCT value of 0.20, corresponding to 65% sensitivity and 45% specificity, when the AUC was 0.63. However, after categorizing this variable, no effect on DFS (p = 0.378) or OS (p = 0.655) was found. Although this biomarker is associated with relapse to some extent, its distribution shows a large overlap between relapsed and non-relapsed patients.

[0098] Dermatopontin (DPT)

[0099] DPT was included in the validation set because its differential expression of relapse in patients outside the Milan criteria was nearly significant (p = 0.06). DPT was strongly associated with relapse rate in the validation set (OR 1.283, CI 1.103–1.494, p = 0.001). The association with relapse rate (HR 1.048, CI 0.977–1.123, p = 0.192) or overall survival (OS) (HR 1.036, CI 0.961–1.116, p = 0.357) was initially not significant. ROC curve analysis showed an AUC of 0.77 (…). Figure 2 The deltaCT cutoff value was set at 7, corresponding to 84% sensitivity and 63% specificity. Using this value, the HR for DFS was observed to be 0.480 (CI 0.295–0.782, p = 0.003). The HR for OR was 0.503 (CI 0.305–0.831, p = 0.007). The strong expression of DPT (OR 0.116, CI 0.037–0.363, p < 0.001) resulted in a nearly 9-fold reduction in recurrence risk.

[0100] In summary, DPT overexpression indicates a favorable prognosis after liver-endocrine leukemia (LT). Patients with strong DPT expression had a 5-fold reduced risk of recurrence, a 50% increased disease-free survival, and a 40% increased overall survival. DPT expression can distinguish patients with a good prognosis from those with a poor prognosis after LT. It was found to be associated with microvascular invasion, possibly synergistically with CLU expression. This association may be related to its function in extracellular matrix assembly and cell adhesion, ultimately affecting microvascular invasion. This is the first report to associate DPT expression with the prognosis of LT due to hepatocellular carcinoma (HCC), and the first demonstration that this gene is an independent predictor of a favorable prognosis.

[0101] Sprouty RTK signaling antagonist 2 (SPRY2)

[0102] This protein is an important regulator of key pathways associated with cancer development, such as angiogenesis, cell growth, invasion, migration, and cytokinesis. Song et al. (Hepatobiliary & Pancreatic Diseases International 2012; 11(2):177-184) used samples from 240 randomly selected HCC patients who underwent hepatectomy to investigate SPRY2 expression on tissue microarrays. Downregulation of SPRY2 expression was observed in 207 patients (86.3%), which was associated with lower survival (p = 0.002) and higher recurrence rates (p = 0.003), serving as an independent predictor of postoperative recurrence in HCC patients (HR = 1.47; 95% CI, 1.02–2.08; p = 0.037). This study is the first to associate SPRY2 overexpression with increased recurrence rates in a patient population undergoing LT treatment for HCC. However, due to significant overlap between the two groups, the inventors were unable to determine a cutoff value that could effectively distinguish between recurrent and non-recurrent patients. Therefore, the clinical relevance of SPRY2 still needs to be verified by future studies.

[0103] SPRY2 showed some differential expression in the trial set for recurrence within the Milan criteria (p = 0.13). In the validation set, SPRY2 was significantly associated with recurrence, with an OR of 1.342 (CI 1.106–1.628, p = 0.003). However, no association was found in DFS (HR 1.050, CI 0.962–1.145, p = 0.273) or OS (HR 1.030, CI 0.943–1.125, p = 0.514). With an AUC of 0.68, we determined the optimal cutoff value to be a deltaCT value of 7.6, corresponding to 69% sensitivity and 65% specificity. However, no correlation was obtained after classification, neither in DFS (HR 1.420, CI 0.897-2.248, p = 0.134) nor in OS (HR 1.305, CI 0.814-2.094, p = 0.269).

[0104] Cluster protein

[0105] In the trial set, CLU revealed differential expression in patients with relapse outside the Milan criteria (P = 0.02), and similarly in the whole set (p = 0.01). According to the Wilcoxon test, CLU expression appeared unrelated to relapse in the validation set. However, logistic regression analysis showed that CLU expression was associated with relapse in the validation set (OR 1.219, CI 1.059–1.403, p = 0.006). These results are not incompatible, as the Wilcoxon test, while a more robust statistical test, is not sensitive to fairly subtle differences between groups; in our case, these differences translated into subtle differences in expression between the relapse and non-relapse groups. Initially, no association was found with DFS (HR 1.073, CI 0.978–1.178, p = 0.136) or OS (HR 1.014, CI 0.913–1.125, p = 0.796). Using ROC curves, the critical value for ACT was determined to be -0.54, corresponding to an AUC of 0.59 (sensitivity 7%, specificity 35%). With this critical value, the HR for DFS was 1.568 (CI 0.951–2.584, p = 0.078), and the HR for OS was 1.550 (CI 0.926–2.595, p = 0.096). Although this association appears weak, we consider it sufficient to be tested in multiple logistic regression and Cox regression analyses. (F-box and WD repeat protein 7 (FBXW7))

[0106] Based on relapse rates in patients within the Milan criteria in the trial set, FBXW7 levels appeared to show differential expression (p = 0.04). However, this difference was not confirmed in the validation set. Logistic regression showed an odds ratio (OR) of 1.238 for relapse rate (CI 0.894–1.052, p = 0.545). Cox regression yielded a hazard ratio (HR) of 1.023 for DFS (CI 0.929–1.128, p = 0.640) and an HR of 1.000 for OS (CI 0.900–1.113, p = 0.994).

[0107] Example 5: Total Population Analysis

[0108] The inventors analyzed the entire population consisting of 180 patients. Taking into account results obtained in the validation set, this analysis included CLU and DPT expression levels measured by ACT values ​​associated with housekeeping gene expression in tumor tissue.

[0109] Factors associated with recurrence were tested using multivariate logistic regression in this population (Table 3). The inventors observed that tumor number, microvascular invasion, and poor differentiation were independently associated with recurrence, as was total histological tumor volume (TTV). Using DPT expression reduced the predicted risk of recurrence by more than 5-fold, while CLU expression predicted a 61% reduction. However, when both were included in the model, only DPT expression still independently predicted recurrence (OR 0.178, CI 0.063–0.507, P = 0.001), not CLU (OR 0.729, CI 0.063–0.507, p = 0.554). No statistical interaction was detected between the variables “DPT expression” and “CLU expression” when recurrence rate was used as the outcome variable (p = 0.402). Expression of both CLU (p = 0.94) and DPT (p = 0.960) was not associated with poorly differentiated tumors.

[0110]

[0111] Table 3: Multivariate logistic regression analysis of relapse in the total population (N=180). OR: odds ratio; Cl: confidence level.

[0112] Strong DPT expression was associated with reduced microvascular invasion (OR 0.370, CI 0.140–0.979, p = 0.045), while CLU expression was not (p = 0.173). However, when microvascular invasion was the outcome variable, an interaction between DPT and CLU was detected (p = 0.033), suggesting that the occurrence of microvascular invasion in tumors with low or high DPT expression is closely related to CLU expression. In this study, 21.2% (14 / 66) of patients with low DPT expression had microvascular invasion. In this subgroup, the incidence of microvascular invasion was 4.8% (1 / 21) with strong CLU expression and 28.9% (13 / 45) with low CLU expression (p = 0.027). This interaction has never been found in the scientific literature and remains to be explained in future studies, suggesting a link between these genes.

[0113]

[0114]

[0115] Table 4: Multivariate Cox regression analysis of disease-free survival rate among 180 patients.

[0116] Table 4 shows the results of multivariate Cox regression analysis of DFS in the total population. Microvascular invasion was the only factor associated with decreased DFS. Conversely, strong expression levels of DPT and CLU were used as positive predictors and were associated with increased survival. Although CLU was no longer an independent variable when both CLU and DPT were included in the model (HR 0.600, CI 0.351–1.026, p = 0.062), no statistical interaction was detected between CLU and DPT (p = 0.822).

[0117] Dermatopontin (DPT)

[0118] The inventors observed DFS based on strong and weak DPT expression. Figure 3 Patients with strong DPT expression (defined as ACT levels above 7) had a disease-free survival (DFS) of 70% and 52.2% at 5 and 10 years, respectively. Strong DPT expression can also be used to identify a subgroup of patients with better prognosis, even in patients outside the MC (Figure 4). This also applies to patients exhibiting other poor prognostic criteria, such as TTV > 115 cm. 3 Patients with microvascular invasion or poor differentiation ( Figure 5A (B and C). Finally, the 5-year and 10-year DFS rates for patients in the HCV subgroup with strong DPT expression were 79.2% and 58.8%, respectively. Figure 6 ).

[0119] Cluster protein (CLU)

[0120] CLU expression in 198 HCC samples was analyzed using tissue microarrays (Wang et al., Oncotarget 2015; 6(5); 2903-16). CLU protein was mainly detected in the cytoplasm of tumor cells. Multivariate Cox regression analysis showed that CLU overexpression was an independent prognostic factor for recurrence after tumor resection (HR 1.628). In addition, the same study found that CLU overexpression significantly promoted the invasion of HCC cells in vitro and promoted distant lung metastasis in vivo, while silencing CLU reduced the invasive ability of HCC cells in vitro and in vivo. CLU overexpression can enhance the metastatic potential of prostate cancer, renal cell carcinoma, gallbladder cancer and breast cancer.

[0121] In the current cohort, CLU overexpression was associated with relapse rate and survival, and was able to distinguish patients with a good prognosis after LT from those with a poor prognosis. CLU also has a significant impact on disease-free survival (DFS). Figure 7 Patients with strong CLU expression had 5-year and 10-year survival rates of 68.9% and 56.1%, respectively. Although CLU expression values ​​were not as significant as DPT, they were effective in selecting patients with better outcomes in the poor prognosis group, such as those outside the Milan criteria (A) or those with TTV > 115 cm. 3 Patients (B) with microvascular invasion (C) or poor differentiation (D) were not statistically significantly associated with survival based on CLU expression, although a clear trend was noted (Figure 8). The 5-year and 10-year DFS rates for the HCV subgroup with strong CLU expression were 74% and 67%, respectively. Figure 9 ).

[0122]

[0123] Table 5: Multivariate Cox regression of overall survival rate.

[0124] Multivariate Cox regression analysis was also performed on overall survival (OS). Ethanol consumption and microvascular invasion were independently associated with decreased OS. Although microvascular invasion is a known risk factor, ethanol consumption led to a decrease in OS without a corresponding decrease in disease-free survival (DFS), which may be related to common comorbidities in ethanol poisoning, such as ethanol-induced cardiomyopathy. Strong expression of DPT and CLU was independently associated with increased OS (Table 5).

[0125] Multivariate combinations with predictive capabilities

[0126] To enhance the prognostic ability of genes associated with HHC outcomes, a multigene model combining the predictive power of all genes was developed to investigate the synergistic effect between DPT and CLU. A simple score based on the expression of the two gene combinations was used. Since their DFS hazard ratios were similar, each gene was assigned the same score: 0 points for weak expression of both genes, 1 point for strong expression of one gene, and 2 points for strong expression of both genes. Strong expression of the two genes was associated with 78% and 60% of 5-year and 10-year survival, respectively, making the gene combination superior to either DPT or CLU alone. Figure 10 This score can differentiate patients outside the Milan criteria. Figure 11A TTV>115cm 3 Patients ( Figure 11B ) and patients with poorly differentiated tumors ( Figure 11C Good prognosis after LT. Similarly, an increasing trend in survival was observed in patients with microvascular invasion, although it did not reach statistical significance. Figure 11D The results were superior to those obtained for each gene individually. Figure 12 This demonstrates the use of this simple combination of gene scores and TTV <115cm 3 This was combined with the predicted survival rates. The combination identified a group of patients with a very good prognosis after LT, a group of patients with an intermediate but still acceptable prognosis after LT, and a group of patients with a poor prognosis after LT.

[0127] Based on previous results for each individual gene, DPT may have better predictive power than CLU. Therefore, combinations of these genes were evaluated without using scoring to better assess the prognostic potential of patients who strongly express only one of the two genes. Figure 13 ).

[0128] For the second approach, which incorporates multiple genes into the HCC prediction strategy, the expression levels of two genes, CLU and DPT, are used along with TTV in a linear kernel support vector machine classification algorithm. To evaluate the algorithm's performance, a Jackknifing strategy is employed: the algorithm is trained using all but one data point, evaluated using that removed data point, and then the process is repeated for each data point, collecting the results. After removing the test sample, the remaining dataset undergoes data augmentation (SMOTE) and normalization (unit variance and zero mean). This is done to obtain a balanced training dataset, prevent overfitting to the most prevalent classification, and ensure that all features are comparable in the classification feature space.

[0129] A cohort of 154 patients was evaluated using a multivariate algorithm with three variables. Variables were binarized using a thresholding technique, where patients below the threshold were assigned a value of 0 for that feature, and patients above the threshold were assigned a value of 1. The threshold units were the same as those for the corresponding features (ACt for gene expression and volume units for tumor volume). The threshold used for CLU was -0.54 ACt, for DPT it was 7 ACt, and for TTV it was 115 cm. 3 Following the evaluation procedure, the following results were obtained, demonstrating that an algorithm combining binary data derived from multiple indicator genes and tumor volume can accurately classify patients at risk of disease recurrence after tumor thrombosis (LT) for HCC treatment (accuracy 67%, false positive rate 22%, precision 91%, recall 64%). Figure 14 , 15 For a subset of patients classified within the Milan criteria used in current clinical practice (accuracy 69%, false positive rate 36%, precision 92%, recall 71%), Figure 16 A subset of patients outside of this subset (accuracy 62%, false positive rate 11%, precision 87%, recall 45%) Figure 17 When classifying patients, the accuracy of classification is also favorable, indicating that classifying patients according to a binary score based on indicator gene expression thresholds and tumor volume can identify a subset of patients who are likely to achieve favorable results after LT, while other patients will be excluded from the procedure.

[0130] For example, for patient 1 with values ​​DPT = 10.48, CLU = 5.34, and TTV = 18.0, and patient 2 with values ​​DPT = 4.39, CLU = 0.84, and TTV = 5.0, these values ​​are binarized using the corresponding thresholds, generating binary values ​​for each variable. These variables are fed into a linear kernel SVM algorithm to produce binary values ​​indicating the predicted outcome of relapse. The model is capable of producing estimated probabilities; in this embodiment, the results themselves are binarized to separate values ​​above and below 50% probability, thus simplifying the output. The interpretation of the results is straightforward, as the output in binary value form only indicates whether a relapse is predicted (R, 0) or whether a relapse is foreseeable (NR, 1) (Table 7).

[0131]

[0132] Table 7: Examples of using the SVM algorithm to classify LT outcomes in HHC patients.

[0133] Finally, heterogeneity was also considered. Samples from different sites of the same tumor were obtained from seven patients. Measurements of changes in expression levels were correlated with the housekeeping genes for CLU and DPT, as well as with whether different values ​​would lead to migration to a group (“group migration”) set by previously calculated thresholds. In these samples, DPT expression levels showed little change, and no patients had to migrate to another group based on these values. On the other hand, CLU expression appeared to be associated with tumor heterogeneity. These results suggest that intratumoral gene expression of DPT is more likely to be homogeneous.

[0134] CAPNS1

[0135] In a study involving 192 patients who underwent total liver transplantation (LT) for HCC, CAPNS1 overexpression was significantly associated with tumor number and size, tumor encapsulation, venous invasion, and pTNM stage. Multivariate analysis showed that CAPN4 expression was a strong independent prognostic factor affecting survival in HCC patients (HR 4.068, CI 2.524–6.555; p < 0.001). The inventors demonstrated a correlation between CAPNS1 expression and recurrence rate (p < 0.001). However, no association with overall survival (OS) or disease-free survival (DFS) was demonstrated, nor was a cutoff value for ACt used to differentiate between patients who will or will not relapse. CAPNS1 is a promising biomarker for combination with other discriminative markers.

[0136] Example 6: Materials and Methods

[0137] Molecular prognostic biomarkers

[0138] Once the clinical factors associated with prognosis were identified, the inventors aimed to determine molecular biomarkers with higher prognostic value for HCC. Therefore, the CHBPT at Curry Cabral Hospital collaborated with Ophiomics-Precision Medicine on a study led by Professor José Pereira Leal, with co-investigator Joana Cardoso Vaz. The molecular biomarker study involved data mining and bioinformatics analysis of public repositories on HCC patients and published literature on HCC biomarkers.

[0139] Data mining and bioinformatics analysis

[0140] Molecular data from HCC patients who underwent liver resection and transplantation were mined and analyzed using public repositories. Available public data on HCC patients, including miRNA and mRNA expression profiles, were reanalyzed using the inventors' analytical workflow.

[0141] Prognostic molecular markers for hepatocellular carcinoma: a systematic review

[0142] The inventors searched the literature for previously published biomarkers in HCC patients who underwent liver resection (LT). Due to the scarcity of data on LT patients and the importance of obtaining information from cohorts of patients undergoing hepatectomy, studies on LR biomarkers were included in the analysis. This paper provides a systematic review of the existing evidence regarding the role of molecular biomarkers in the prognosis of patients undergoing HCC resection or transplantation. This review followed the general guidelines of the Institute of Medicine, National Academies of Sciences' Systematic Review Criteria. The aim of this review was to assess the role of molecular biomarkers in the prognosis of patients treated with LR or LT for HCC.

[0143] Overall: The search was limited to articles in English-language journals from January 2008 to October 2016. Studies published only as abstracts, unpublished studies, and articles published in non-peer-reviewed journals were excluded. Animal and in vitro studies were also excluded. Studies of molecular markers unrelated to HCC prognosis were also excluded. Intervention: Liver resection and transplantation in patients with HCC. The reason for choosing resection was the lack of research on LT. Outcomes: The primary outcomes of the search were disease-free survival, recurrence rate, and overall survival. Study Design: Due to the rarity of randomized controlled trials (RCTs) and controlled trials in this case, and due to the scarcity of data, cohort studies (even retrospective studies) were eligible for inclusion. Case series were accepted, while case reports were excluded from evaluation. Economic assessments were also excluded. PubMed, Clinicalkey, and Cochrane databases were searched using different combinations of the following keywords: hepatocellular carcinoma, surgery, resection, transplantation, prognosis, molecular and biomarkers. References cited in the identified articles were used to find further relevant publications. The retrieved data included data types (mRNA, miRNA, and protein), prognostic information, specific genes involved, genes indicating good or bad prognosis, alteration types (overexpression, downregulation, hyper / hypomethylation, and mutation), patient samples, statistical data, and author observations. Biomarkers were selected based on predictive power, citation counts across different centers, replication technology capabilities, and available reagents. Preliminary studies were conducted after rigorous screening to identify the best candidate biomarkers.

[0144] Study population

[0145] Inclusion and exclusion criteria were the same as those used in previous clinical biomarker studies. Inclusion criteria: Patients who underwent liver transplantation for hepatocellular carcinoma. Exclusion criteria: Age under 18 years; no cirrhosis; fibrous lamellar histological type or cholangiocarcinoma histological type; lack of histological evidence of HCC; furthermore, since the inventors' primary outcome measure was recurrence rate / disease-free survival, cases with perioperative mortality, extrahepatic invasion, and residual disease were excluded. A complete population of 301 patients who underwent liver transplantation for HCC between September 1992 and February 2014 was considered. Of the 231 patients obtained after applying the exclusion criteria, the inventors included only those with a follow-up period of more than 5 years. The final sample set used in this study (trial set and validation set) included 180 patients. Trial set: The trial set included patients outside the Milan criteria (n=6) and those without recurrence (n=7), and patients with early recurrence within the Milan criteria (n=6) and those without recurrence (n=7). A subset of biomarkers that met the selection criteria in the trial study underwent a second round of analysis via the validation set. Validation set: By excluding patients previously included in the trial set, patients who died perioperatively, patients with extrahepatic invasion, patients with residual disease, and patients who were followed up for less than 5 years from the initial 275 patients, the initial validation set consisted of 154 patients.

[0146] Sample collection

[0147] Tumor specimens from 180 patients who underwent liver transplantation at the Hepatobiliary and Pancreatic Transplant Center of Curry Cabral Hospital were fixed in formalin and preserved in paraffin blocks. For all selected blocks, histopathological characterization and regional selection were performed on hematoxylin and eosin (HE) stained sections under the supervision of experienced pathologists. This study was approved by the ethics review committees of Curry Cabral Hospital and NOVA School of Medicine.

[0148] RNA extraction and cDNA synthesis

[0149] Archived FFPE tissue sections (5 μm) were dewaxed and counterstained with Mayer's hematoxylin and eosin. All samples underwent macroscopic dissection under the guidance of a pathologist. Total RNA was extracted using the RNeasyFFPE kit (Qiagen) with slight modifications to the manufacturer's instructions: proteinase K cell lysis was performed overnight at 56°C. The RNase-FreeDNase Set (Qiagen) "on column" DNA digestion program was included. SuperScript™ VILO was used. TM The cDNA synthesis kit (Thermo Fisher Scientific) performs reverse transcription on each extracted RNA sample.

[0150]

[0151]

[0152] Table 6: Primer sequences for target and reference genes.

[0153] Quantitative Real-Time PCR

[0154] Due to known FFPE degradation issues and the limited sample size, standard methods cannot be used to assess RNA concentration and integrity. Therefore, the inventors have used a high-sensitivity RNA ScreenTape (Agilent) on an Agilent 2200 TapeStation system to determine the quantity and quality of isolated RNA samples. Primer sets were designed using the NCBI Primer-BLAST tool (Ye et al, BMC Bioinformatics 2012; 13:134), operating at 60°C, with amplicon length of 70-100 bp (Table 6), and were purchased from Invitrogen (Thermo Fisher Scientific). In 10 μL of a reaction mixture containing template (1 μL, 0.5-1 ng / μL) and primers (0.5 μM each), SsoFast was used. TM Supermix (Bio-Rad, Hercules CA, USA) reagents were used to analyze the replicas of each sample via RT-qPCR. The replication was performed on a CFX96 Touch according to the cycle plan. TM Samples were processed in a real-time PCR detection system (Bio-Rad, Hercules CA, USA) for 50 cycles: 98°C for 120 s, 98°C for 5 s, and 60°C for 15 s. Fluorescence data were collected at 60°C. Data and statistical analysis were performed. The relative differential expression analysis of the target gene by RT-qPCR was based on the 2-AACt or ACt method to calculate the fold change in expression, where a value greater than 1 indicates upregulation and a value less than 1 indicates downregulation (Livak et al, Methods 2001; 25(4):402-8), and the average quantification period of the replicate was used as the period threshold (Ct) for comparison with the Ct of the calibration gene ribosomal protein L13a (RPL13A). Differential expression of target genes using RT-qPCR data between a sample set with disease relapse (R) and a sample set without disease relapse (nonR) was statistically computed using R (Team RDC, Austria, 2009) and statistical significance was calculated using the Wilcoxon rank-sum test (confidence level = 0.95).

[0155] Furthermore, using the same methods as previously described for clinical biomarkers, the obtained RT-qPCR data were correlated with disease-free survival for each candidate target gene. Continuous variables were expressed as median, i.e., interquartile range (IQR), or mean and standard deviation (SD), and compared using independent samples t-tests. Demographic variables of the target transplant patients were compared using Student's t-test, Pearson's chi-square test, or Fisher's exact test, depending on the circumstances. Outcome variables were relapse (disease-free survival) and death (overall survival). The time to outcome was calculated using the transplant date up to the date of the event or the last follow-up date for patients who did not experience the event. Kaplan-Meier survival curves were constructed for post-transplant outcome analysis. The effects of demographic variables on disease-free survival and overall survival were investigated using log-rank tests and Cox regression models. Criterion values ​​were determined by receiver operating characteristic (ROC) analysis. For multivariate analysis, all variables significant for outcomes (P < 0.20) were included in a Cox proportional hazards model (Therneau et al.; Springer-Verlag, 2000) or a type-dependent multivariate logistic regression model. Backward selection was performed to preserve significant variables. Separate stratified survival analyses were performed as necessary. A p-value less than 0.05 was considered significant. Statistical analyses were performed using SPSS versions 22.0 and 24.0 (SPSS Inc., Chicago, IL). The linear kernel SVM algorithm was developed in a Python 3.6.7 environment, using the following packages for data manipulation and classification: scikit-learn (0.23.2); numpy (1.19.1); pandas (0.23.4). SEQUENCE LISTING <110> Ofomix Biotechnology Research and Development AG <120> Clinical and molecular prognostic biomarkers for liver transplantation <130> IP2204293PT <140> 2020800837166 <141> 2020-10-02 <150> 19201218.5 <151> 2019-10-02 <160> 34 <170> PatentIn version 3.5 <210> 1 <211> 19 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 1 ctggaagcac cggaaggag 19 <210> 2 <211> 22 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 2 ccatatgttc tccagagcgc ag 22 <210> 3 <211> 24 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 3 attccaggtc tatggaaatg cagg 24 <210> 4 <211> 23 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 4 cgtcagatcc ccaaaaagaa aat 23 <210> 5 <211> 18 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 5 gcaacaacga cgccgaat 18 <210> 6 <211> 18 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 6 ccgcgatcac ggagttca 18 <210> 7 <211> 17 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 7 aactacgcct gcatgcc 17 <210> 8 <211> 17 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 8 ctacgtgccc tgcagcc 17 <210> 9 <211> 18 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 9 gcctgtgagc tgaagggg 18 <210> 10 <211> 20 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 10 actctgctgc tgtttgtggg 20 <210> 11 <211> 23 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 11 gactcacaca ctggagagaa gcc 23 <210> 12 <211> 20 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 12 gttgtgacac gacaccctga 20 <210> 13 <211> 20 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 13 gggacatggg cagagcaatg 20 <210> 14 <211> 18 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 14 cacccaggct ctgcagtc 18 <210> 15 <211> 21 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 15 gatgctcagt gaggaccctt g 21 <210> 16 <211> 21 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 16 cctaccttc acctggtagc c 21 <210> 17 <211> 19 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 17 ccttcgatgg agtgccagg 19 <210> 18 <211> 23 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 18 gtaaatgtcc gatttcgtga gcc 23 <210> 19 <211> 23 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 19 cagtttggag ctacctggag tat 23 <210> 20 <211> 22 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 20 aagtagccca caagagtaag ca 22 <210> 21 <211> 20 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 21 gtccactcca gctctgaaac 20 <210> 22 <211> 19 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 22 ccactctgag ctctggcct 19 <210> 23 <211> 20 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 23 gcacgtctgg taccattcca 20 <210> 24 <211> 16 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 24 ccggtgaagc ctggca 16 <210> 25 <211> 23 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 25 cattgattcc ccgcttcatt gtg 23 <210> 26 <211> 22 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 26 tggacatttc ctggagctca tt 22 <210> 27 <211> 23 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 27 catgtagtga acctttaagt tgc 23 <210> 28 <211> 17 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 28 atccatcacg gccacca 17 <210> 29 <211> 22 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 29 ataaatctgc gtgggtagga gc 22 <210> 30 <211> 20 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 30 gggtacggct ggtagttgtc 20 <210> 31 <211> 19 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 31 ggtgacatcg accacctcc 19 <210> 32 <211> 24 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 32 gtctcaagga tacctgctac agaa 24 <210> 33 <211> 20 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 33 cgtgcgaggt atgctgcccc 20 <210> 34 <211> 20 <212> DNA <213> Artificial Sequence <220> <223> Primer <400> 34 ggggtggga tgccgtcaaa 20

Claims

1. Use of a reagent for detecting the expression level of a biomarker combination in the preparation of a product for predicting and / or diagnosing the prognosis of hepatocellular carcinoma after liver transplantation, characterized in that: The biomarker composition is an indicator gene; The indicator genes include: DPT, gene ID GC01M168664.

2. The use according to claim 1, characterized in that The indicator genes also include: CLU, gene ID GC08M027596.

3. The use according to any one of claim 2, characterized in that: Overexpression of the indicator gene relative to a threshold value indicates a good prognosis.

4. The use according to claim 3, characterized in that: Overexpression of DPT in the indicator gene indicates a good prognosis; or Overexpression of DPT and CLU among the indicator genes indicates a good prognosis.

5. The use according to claim 4, characterized in that: Overexpression of the indicator genes DPT and CLU and total tumor volume ≤ 115 cm 3 Indicates a good prognosis.

6. The use according to claim 3, characterized in that: The indicator gene expression level is determined by polymerase chain reaction; wherein the indicator gene expression value relative to the threshold is determined as the difference between the threshold cycle number of the indicator gene and the threshold cycle number of the internal control gene; The threshold cycle number refers to the PCR cycle number when the product is detected.

7. The use according to claim 6, characterized in that: If the difference between the threshold cycle numbers of the indicator gene is as follows, the indicator gene is considered to be overexpressed: DPT greater than 7; and CLU is higher than -0.

54.

8. The use according to any one of claims 1 or 2, characterized in that: The expression levels of indicator genes are incorporated into the algorithm to provide a value reflecting the likelihood of disease recurrence.

9. The use according to any one of claims 1 or 2, characterized in that: The expression levels of indicator genes and tumor volume measurements were incorporated into the algorithm to provide a value reflecting the likelihood of disease recurrence.

Images