Non-functional pancreatic neuroendocrine tumor prognostic markers and uses thereof

Abstract

The application provides a non-functional pancreatic neuroendocrine tumor prognostic marker and application thereof, the prognostic marker is one or more of GNAO1, INA and VCAN;Wherein GNAO1 and INA high expression is related to good prognosis, and VCAN high expression is related to poor prognosis.Three prognostic related proteins are combined into a multivariate prognostic marker, the multivariate prognostic marker is excellent in distinguishing high-risk patients and low-risk patients, and shows good calibration effect in estimating absolute risk, and the prediction error rate is low.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of biological medicine, and particularly relates to a prognostic marker for non-functional pancreatic neuroendocrine tumor and application thereof. BACKGROUND

[0002] Pancreatic neuroendocrine tumors (PanNENs) are the second most common epithelial tumors of the pancreas, with a significant increase in incidence over the past four decades. According to the World Health Organization (WHO) classification, pancreatic neuroendocrine tumors can be divided into well-differentiated pancreatic neuroendocrine tumors (PanNETs) and poorly differentiated pancreatic neuroendocrine carcinomas (PanNECs). Among them, PanNETs account for more than 90% of all pancreatic neuroendocrine tumors, and are further divided into functional PanNETs and non-functional pancreatic neuroendocrine tumors (NF-PanNETs) according to whether there is a clinical hormone hypersecretion syndrome. NF-PanNETs account for about 90% of PanNETs, and their clinical manifestations are highly diverse. Although most NF-PanNETs are clinically inert, about half of the patients have distant metastasis (mostly liver metastasis) at the time of diagnosis, with an average survival period of only 23 months, while the average survival period of patients with local lesions can reach 124 months. In clinical practice, for high-risk PanNETs patients, it is recommended to conduct imaging monitoring every 3 to 6 months after radical resection, and adjuvant therapy can be considered within the framework of clinical trials. However, there is no consensus on the best guidelines at present, and the over-treatment of NF-PanNETs has been controversial in recent years, and some patients may be more suitable for an observational treatment strategy. Therefore, it is particularly urgent to develop a prognostic marker that can accurately stratify the risk of NF-PanNETs patients, so as to optimize the treatment options and disease management of patients.

[0003] At present, the prognosis prediction of NF-PanNETs mainly relies on clinical pathological variables, such as tumor size, World Health Organization (WHO) grade classification, and lymph node metastasis status (N stage), and these clinical prognostic factors have obvious limitations in terms of prediction performance and application. According to the guidelines of the National Comprehensive Cancer Network (NCCN), the European Neuroendocrine Tumor Society (ENETS), and the North American Neuroendocrine Tumor Society (NANETS), for NF-PanNETs with a diameter greater than 2.0 centimeters, surgical resection is recommended, and a negative surgical margin and regional lymph node dissection are pursued. For NF-PanNETs with a diameter of ≤2.0 centimeters, since they usually do not metastasize, selective surgical resection or observation alone can be considered. However, some NF-PanNETs with a diameter of ≤2.0 centimeters may still exhibit invasive behavior, and even some tumors classified as inert may metastasize.

[0004] In addition, a small number of studies have also begun to explore markers based on traditional molecular detection techniques (such as fluorescence in situ hybridization and immunohistochemistry), including alternative lengthening of telomeres (ALT) status, DAXX / ATRX expression levels, and ARX / PDX1 expression levels, but published studies have shown inconsistencies in the relevant conclusions. For example, Jiao et al. found that patients with DAXX / ATRX expression loss had improved overall survival compared to wild-type NF-PanNETs patients; however, other studies have shown that DAXX / ATRX expression loss is associated with tumor metastasis, shorter disease-free survival (DFS), and shorter disease-specific survival (DSS). This inconsistency may be due to differences in study populations, small sample sizes, and the limitations of single-center studies. These potential biomarkers have not been fully evaluated, and there is a lack of large-scale internal and multi-center external cohort validation. Therefore, there is an urgent need to develop higher-resolution biomarkers by combining multi-omic molecular data to improve the prognostic classification and management of NF-PanNETs patients.

[0005] With the rapid development of high-throughput technologies such as second-generation sequencing and high-resolution mass spectrometry, multi-omic studies such as genomics, transcriptomics, and proteomics have been widely conducted in the field of oncology and have been applied to the development of molecular markers for tumor patient prognosis (such as disease recurrence, progression, and death). In addition, studies have found that compared to genetic expression variations, protein expression and post-translational modifications can better reflect pathological and physiological changes in tumors, which is of great significance for the development of molecular markers that are more closely related to tumor clinical manifestations.

[0006] Based on high-precision proteomic data from a clinical cohort of 108 NF-PanNETs patients, the inventors developed a three-protein prognostic marker using a self-developed ReProMSig analysis platform, which can effectively distinguish between high-risk and low-risk NF-PanNETs patients. The prognostic value of the marker and the constituent proteins was widely validated in multiple internal and external independent cohorts (a total of 500 samples). In addition, using pancreatic cancer cell lines and cell line-derived xenograft (CDX) models, the inventors further confirmed the effects of each marker protein on tumor cell proliferation in vitro and in vivo. SUMMARY

[0007] To overcome the problems of the prior art, the present application aims to identify high-resolution biomarkers related to the prognosis of NF-PanNETs and to identify high-risk (poor prognosis) NF-PanNETs populations, thereby providing a non-functional pancreatic neuroendocrine tumor prognostic marker and its application.

[0008] The object of the present application is achieved in that:

[0009] The first aspect of the present application provides a non-functional pancreatic neuroendocrine tumor prognostic marker, wherein the prognostic marker is one or more of GNAO1, INA and VCAN.

[0010] Further, the prognostic marker is RNA, or reverse-transcribed cDNA thereof, or protein.

[0011] Further, high expression of GNAO1 and INA is associated with good prognosis, and high expression of VCAN is associated with poor prognosis.

[0012] Further, the three prognostic-related proteins: GNAO1, INA and VCAN are combined into a multivariate prognostic marker, and a prognostic risk score is calculated to determine the prognosis of the patient:

[0013] The second aspect of the present application provides use of a reagent for detecting a prognostic marker in the preparation of a product for evaluating the prognosis of a non-functional pancreatic neuroendocrine tumor, wherein the prognostic marker is one or more of GNAO1, INA and VCAN.

[0014] Further, the sample to be detected is a non-functional pancreatic neuroendocrine tumor tissue sample, and the reagent is any one of protein profiling, transcriptome sequencing, gene chip, NanoString, PCR, and immunohistochemistry-related reagent.

[0015] The third aspect of the present application provides a non-functional pancreatic neuroendocrine tumor prognostic risk evaluation kit, wherein the kit comprises a quantitative reagent for quantitatively detecting the expression level of the prognostic marker according to the first aspect.

[0016] The fourth aspect of the present application provides use of the prognostic marker according to the first aspect in risk stratification and precision treatment research of non-functional pancreatic neuroendocrine tumor patients for non-therapeutic purposes.

[0017] The fifth aspect of the present application provides a non-functional pancreatic neuroendocrine tumor patient prognostic risk analysis system using the prognostic marker according to the first aspect, wherein the system comprises:

[0018] A target expression level detection device for detecting the expression level of the prognostic marker according to the first aspect in a sample; the sample is a tumor tissue of a patient;

[0019] A prognostic risk analysis device for calculating a prognostic risk score based on the expression level of the prognostic marker, thereby determining the prognosis of the patient:

[0020] Result output device: for outputting the results analyzed by the prognosis risk analysis device, the prognosis risk analysis including predicting the possibility of the patient developing adverse events (such as tumor recurrence and death, etc.).

[0021] In some embodiments, when the data type applied by the marker expression is mass spectrometry (MS) quantitative data, the prognosis risk score is calculated by the following formula:

[0022] Risk score = -0.504 * GNAO1 - 0.499 * INA + 0.32 * VCAN.

[0023] In addition, in some embodiments, the data type applied by the marker also includes, but is not limited to, patient immunocytochemistry (IHC) marker quantitative data.

[0024] The sixth aspect of the present application provides the use of a reagent for detecting VCAN (Versican, a multi-functional proteoglycan) in the preparation of an early diagnosis reagent for non-functional pancreatic neuroendocrine tumors.

[0025] The advantages and beneficial effects of the present application are:

[0026] 1. The present application first determines that GNAO1, INA and VCAN, and their combination are closely related to the prognosis of NF-PanNETs. The marker can be used to predict the prognosis risk of NF-PanNETs, i.e. to predict the possible outcome information of cancer patients (e.g. cancer recurrence, progression, death, etc.). By performing mass spectrometry quantitative analysis on the tumor tissue of NF-PanNETs patients, or by detecting the expression of related molecules through immunohistochemistry, RT-qPCR, and whole transcriptome sequencing (RNA-seq), the NF-PanNETs patients can be quantitatively risk stratified, and the detection results show that the prognosis marker described in the present application performs excellently in distinguishing high-risk patients and low-risk patients (5-year AUROC reaches 0.881), and performs well in estimating the absolute risk, with a low prediction error rate. The treatment plan selection of patients can be optimized according to the detection results, and the design of clinical treatment and clinical trials can be assisted.

[0027] 2. The present application determines that the secreted protein VCAN can be used as a non-invasive biomarker for early screening of NF-PanNETs. By performing enzyme-linked immunosorbent assay (ELISA) on the peripheral blood of NF-PanNETs patients, the plasma level of VCAN protein can be detected, and whether NF-PanNETs occurs can be evaluated according to the plasma level of VCAN. BRIEF DESCRIPTION OF DRAWINGS

[0028] The application is further illustrated below with reference to the accompanying drawings and examples.

[0029] Figure 1 Schematic diagram of the development and validation process of the prognostic markers of NF-PanNETs of the application;

[0030] Figure 2 Correlation between the expression levels of the three marker proteins and the risk scores predicted by the markers; wherein, A shows the relative abundance of the marker proteins and the predicted risk scores of each sample in the discovery cohort, and the color depth is proportional to the protein abundance; B shows the immunohistochemical (IHC) staining images of the marker proteins of representative patients in the high-risk group and the low-risk group;

[0031] Figure 3 Evaluation results of the prediction performance of the prognostic markers; wherein, A is the independence test of the markers in the discovery cohort; B is the independence test in the validation cohort; C is the evaluation of the discrimination ability of the markers and their proteins in distinguishing high-risk patients from low-risk patients;

[0032] Figure 4 Comparison results of the prediction performance between the prognostic markers and multiple NF-PanNETs prognostic-related clinicopathological variables, including time-dependent ROC curves for 1 year, 3 years and 5 years;

[0033] Figure 5 Kaplan-Meier survival curves between the high-risk patient group and the low-risk patient group predicted by the prognostic markers in the discovery cohort and the validation cohort, reflecting the difference in survival probability between the two groups; wherein, A is the discovery cohort, and B is the validation cohort (DIA-MS);

[0034] Figure 6 Survival probability difference between the high-risk patient group and the low-risk patient group predicted by the prognostic markers in other validation cohorts (two internal TMA validation cohorts and one external validation cohort);

[0035] Figure 7 Survival probability difference between the high-risk patient group and the low-risk patient group predicted by the prognostic markers in the discovery cohort and multiple internal and external validation cohorts;

[0036] Figure 8 The influence of each marker protein on tumor cell proliferation was verified in vitro and in vivo using pancreatic cancer cell lines and cell line-derived xenotransplant tumor CDX models; wherein, A is the cell line experiment, and B is the CDX model experiment;

[0037] Figure 9The expression levels of the marker protein VCAN in the plasma of normal people and tumor patients, wherein A is the difference in the plasma expression level of VCAN between healthy people and tumor patients, B is the difference in the plasma expression level of VCAN between healthy people and tumor patients in the first stage, and C is the correlation between the plasma expression level of VCAN and the expression amount of VCAN immunohistochemical (IHC) labeling in the corresponding tumor tissue. DETAILED DESCRIPTION

[0038] The examples are used to better illustrate the present application, but the present application is not limited to the examples. Therefore, those skilled in the art can make non-essential improvements and adjustments to the embodiments according to the above disclosure, which still fall within the protection scope of the present application.

[0039] The endpoints of the ranges and any values disclosed herein are not limited to the precise values stated. The endpoints of the ranges and the values are approximate values and should be understood as including values approximately near to the stated values. For ranges of values, the endpoints of the ranges are included in the ranges, and the endpoints and individual points within the ranges are combinable to form new ranges, which are also within the scope of the present disclosure.

[0040] The present application will be described in detail below through examples. It should be understood that the following examples are only used to exemplarily further explain and illustrate the content of the present application, and are not used to limit the present application. Example 1: Screening and determination of NF-PanNETs prognostic markers

[0041] In this example, 144 samples of pancreatic neuroendocrine tumors and their paired normal adjacent tissues (NATs) were collected. After excluding patients with neuroendocrine cancer or functional pancreatic neuroendocrine tumors, and patients who had received neoadjuvant therapy or lacked imaging data during follow-up, 108 cases of untreated NF-PanNETs were finally included as a discovery cohort. TMT-labeled whole proteome mass spectrometry analysis was performed on the discovery cohort and the corresponding normal tissues. In addition, 51 cases of NF-PanNETs were used as an independent verification cohort, and the tumor samples were analyzed by data-independent acquisition mass spectrometry (DIA-MS). The proteomic data of the discovery cohort and the verification cohort were used to develop and verify the multi-protein markers provided by the present application.

[0042] In the discovery cohort, the present embodiment identified 150 proteins that were significantly upregulated in NF-PanNETs tumor samples (compared to normal tissue from the same patient), including: ABCB11, ACTL6B, ADAMTSL2, ANKS4B, ARMCX4, ATP1A3, ATP6AP2, ATP9A, BASP1, BRAWNIN, BSN, CA11, CACNA2D2, CACNB2, CAMK2B, CD200, CDYL2, C ELF3, C ELF4, C ELF5, C ELF6, CERS6, CGREF1, CHGA, CHGB, CKB, CKMT1A, CLGN, COL22A1, COL8A2, CPE, CPLX1, CPLX2, CRIP2, CRMP1, CRYBA2, CTHRC1, DYNC1I1, EEF1A2, ELAVL2, ELOVL5, EN02, ENPP2, FAM155A, FAM169A, FBLL1, FIBIN, GDI1, GNAO1, GNAZ, GNG4, GPRIN1, GREM2, GRIA2, HAPLN1, HDGFL3, HMGB3, HMGN5, IGFBP3, IGFBP7, INA, KHDRBS2, KIF5C, KNDC1, LGALS3BP, LY6H, MAP1B, MAPK8IP1, MARCKSL1, MARK1, MDK, MT3, MUC13, NAPB, NCAM1, NOVA1, OCC1, P4HTM, PAM, PCP4, PCSK1, PCSK1N, PFKP, PKIA, POSTN, PPP1R14C, PRRT1B, PTGDR2, PTMS, PTS, PVR, QPCT, RAB39A, RAB3A, RAB3B, RAB3C, RAI2, RAPGEF5, REEP2, RHBDD2, RIC3, RTN1, RUNDC3A, SCAMP5, SCD, SCG2, SCG3, SCGN, SDHAF4, SEPTIN3, SERPINE2, SFRP4, SH3GL2, SLC12A5, SLC7A14, SLC7A8, SMOC1, SNAP25, SNAP91, SOGA3, SRSF12, SSBP4, SSTR2, STMN3, STX1A, STXBP1, SV2A, SVOP, SYN1, SYP, SYT5, SYT7, TCEAL2, TCEAL3, TCEAL5, TFF3, TM4SF4, TM4SF5, TMEM164, TMEM176B, TNC, TOX, TP PP3, UCHL1, UCN3, VCAN, VGF, VSTM2L, WNT4, XKR7.

[0043] As shown in Figure 1 The present embodiment utilizes bioinformatics analysis and ReProMSig platform to develop and evaluate multi-protein prognostic markers. ReProMSig adopts a bootstrap strategy to screen out robust proteins significantly related to prognosis from the above-mentioned 150 up-regulated proteins of NF-PanNETs. The strategy aims to correct the selection bias of predictive factors, and the specific steps include:

[0044] (i) 200 bootstrap samplings are performed on the training set (i.e. the discovery cohort) to create multiple protein quantification matrices;

[0045] (ii) Supervised principal component analysis (SPCA) is performed on each matrix to identify a set of important predictive factors;

[0046] Finally, it is found that GNAO1, INA and VCAN are selected by more than 75% of the models, so they are determined as the most robust predictive factors related to prognosis and are included in the final LASSO-penalized Cox regression model, obtaining the NF-PanNETs prognostic markers GNAO1, INA and VCAN described in the present application, wherein, as shown in Figure 2 GNAO1 and INA high expression is associated with good prognosis, and VCAN high expression is associated with poor prognosis.

[0047] Taking mass spectrometry data as an example, the above three prognostic related proteins are combined into a multivariate prognostic marker, and the risk score calculation formula is: risk score = -0.504*GNAO1-0.499*INA+0.32*VCAN). By substituting the mass spectrometry quantification value of the three proteins of the patient into the above formula, the risk score of the patient can be calculated.

[0048] Example 2: Evaluation of the prediction performance of NF-PanNETs prognostic markers

[0049] In order to further improve the comprehensiveness of the evaluation of the prognostic markers, the present embodiment includes multiple cohorts to evaluate the prediction performance of the prognostic markers screened in Example 1: including an internal independent validation cohort (DIA-MS, containing 51 tumors); two tissue microarray (TMA) data sets (containing 76 and 98 tumors, respectively, IHC labeling data); an internal independent cohort (containing 108 tumors and 79 healthy controls, ELISA detection data); and an external validation cohort from Shanghai Changhai Hospital (containing 88 tumors). In the discovery cohort and the DIA-MS internal independent validation cohort, through multivariate Cox regression analysis, it is found that the multivariate prognostic marker composed of GNAO1, INA and VCAN is an independent prognostic risk factor Figure 3The prognostic value of the multivariate prognostic marker described in Example 1 was independent of common clinicopathological risk factors and DAXX / ATRX protein expression abundance.

[0050] The multivariate prognostic marker described in Example 1 was validated in the discovery cohort by time-dependent ROC analysis Figure 3 The multivariate prognostic marker described in Example 1 was validated in the discovery cohort by time-dependent ROC analysis

[0051] In addition, the present example directly compared the performance of the present NF-PanNETs prognostic marker and key clinicopathological risk factors (grade, tumor size, TNM stage and perineural invasion) in terms of discriminative ability (i.e. sensitivity and false positive rate). The higher the area under the ROC curve (AUROC), the more accurate the prognostic model. The ROC analysis results showed that the prognostic prediction performance of the present NF-PanNETs marker was superior to all the above-mentioned clinicopathological variables, which was reflected in its prediction of 1-year, 3-year and 5-year survival probability, i.e. with the highest AUROC value (see Figure 4 ; 1-year: 0.846; 3-year 0.844; 5-year 0.881).

[0052] Meanwhile, in the discovery cohort Figure 5 and the DIA-MS internal independent validation cohort Figure 5 , the multivariate prognostic marker described in Example 1 showed significant survival differences between the high-risk and low-risk patients predicted by the marker, with 5-year survival probabilities of 51.4% and 97.8% respectively. The prognostic value of the multivariate prognostic marker and its marker proteins was also verified in other validation cohorts Figure 6 and Figure 7 For high-risk patients predicted by the multivariate prognostic marker, regular postoperative CT follow-up monitoring can be recommended to monitor for recurrence. The effect of each marker protein on tumor cell proliferation was verified in vitro and in vivo using pancreatic cancer cell lines and cell line-derived CDX models, respectively Figure 8 . The expression level of the secreted protein VCAN in the plasma of normal people and tumor patients was determined by ELISA technology, and it was determined that VCAN can be used as a non-invasive biomarker for early screening of NF-PanNETs Figure 9 .

[0053] It should be pointed out finally that the above merely serves to illustrate the technical solutions of the present application but not to limit the present application. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced without departing from the spirit and scope of the present application.

Claims

1. A combination marker for the prognosis evaluation of non-functional pancreatic neuroendocrine tumors, characterized in that, The combined marker is composed of GNAO1, INA and VCAN; wherein high expression of GNAO1 and INA is associated with good prognosis, and high expression of VCAN is associated with poor prognosis.

2. The marker for combination according to claim 1, characterized by, The combined marker is RNA, or reverse-transcribed cDNA thereof, or protein.

3. Use of a reagent for detecting the markers of the combination according to claim 1 or 2, for the preparation of a product for the evaluation of the prognostic situation of non-functional pancreatic neuroendocrine tumors, characterized in that, The combined marker is composed of GNAO1, INA and VCAN.

4. Use according to claim 3, characterized in that, The sample to be detected is a non-functional pancreatic neuroendocrine tumor tissue sample, and the reagent is any one of protein spectrum, transcriptome sequencing, gene chip, NanoString, PCR, and immunohistochemical reagent.

5. A non-functional pancreatic neuroendocrine tumor prognosis risk assessment kit, characterized by, The kit comprises quantitative reagents for quantitatively detecting the expression amount of each marker molecule in the combined marker according to claim 1 or 2.

6. A non-functional pancreatic neuroendocrine tumor patient prognosis risk analysis system using the combination marker according to claim 1 or 2, characterized by, The system comprises: a target point expression amount detection device for detecting the expression amount of each marker molecule in the combined marker according to claim 1 or 2 in a sample; the sample to be detected is a tumor tissue of a patient; a prognosis risk analysis device for calculating a comprehensive prognosis risk score based on the expression amount of each marker molecule in the combined marker, so as to determine the prognosis of the patient; a result output device for outputting the result analyzed by the prognosis risk analysis device.

Images