CN114155956A - System for predicting blood vessel invasion probability of primary liver cancer patient incapable of being resected by surgery - Google Patents
System for predicting blood vessel invasion probability of primary liver cancer patient incapable of being resected by surgery Download PDFInfo
- Publication number
- CN114155956A CN114155956A CN202111455870.4A CN202111455870A CN114155956A CN 114155956 A CN114155956 A CN 114155956A CN 202111455870 A CN202111455870 A CN 202111455870A CN 114155956 A CN114155956 A CN 114155956A
- Authority
- CN
- China
- Prior art keywords
- score
- liver cancer
- probability
- data
- treatment
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Public Health (AREA)
- Medical Informatics (AREA)
- Biomedical Technology (AREA)
- Business, Economics & Management (AREA)
- Pathology (AREA)
- Human Resources & Organizations (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Databases & Information Systems (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Data Mining & Analysis (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Investigating Or Analysing Biological Materials (AREA)
- Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)
Abstract
The invention relates to a system for predicting the blood vessel invasion probability of a primary liver cancer patient which cannot be resected by surgery, which comprises the following components: a data collection module: data for obtaining the patient's etiology, BCLC staging, treatment modality, whether or not cirrhosis of the liver, tumor diameter, alpha-fetoprotein (AFP) levels, and C-reactive protein (CRP) levels; a module for calculating the blood vessel invasion probability: and assigning the data of the etiology, the BCLC stage, the treatment mode, whether the liver is hardened, the tumor diameter, the Alpha Fetoprotein (AFP) level and the C-reactive protein (CRP) level by using the established visual Nomogram model for predicting the vascular invasion of the primary liver cancer patient which cannot be resected by surgery, calculating a total score, and obtaining the corresponding vascular invasion probability according to the total score.
Description
Technical Field
The invention belongs to the technical field of information processing applicable to the medical field, and particularly relates to a system for predicting blood vessel invasion probability of a liver cancer patient.
Background
Primary liver cancer (primary hepatic carcinoma) is one of the common malignant tumors in China. According to 2020-year global cancer burden data published by the international agency for research on cancer (IARC) of the world health organization, the primary liver cancer is the fifth ranked cancer in China according to the number of the onset of the cancer; primary liver cancer ranks second, second only to lung cancer, as calculated by cancer-related mortality. Therefore, primary liver cancer imposes a heavy burden on both patients and medical care systems. Surgical resection, liver transplantation, Transcatheter Arterial Chemoembolization (TACE), radiofrequency ablation (RFA), and chemotherapy are currently the primary means of treating primary liver cancer. These methods can prolong the life of the patient to some extent. However, the recurrence rate of liver cancer is high regardless of the treatment method. It has been reported that approximately 68.9% of patients develop tumor recurrence after RFA (Kocabayoglu P, et al. Expression of fibrous markers in tumor and tumor-bearing tissue at time of transplantation activities with recurrence of hepatic fibrous in tissue intersection transplantation Ann transplantation 2017; 22: 446-54.). Many studies have shown that macrovascular invasion (macrovascular invasion) of portal, hepatic and inferior vena cava is an independent predictor of recurrence and poor prognosis in patients with primary liver cancer. Yao Liu et al reported a scoring model for predicting large vessel invasion by early and middle liver cell carcinoma (Yao Liu et al. A new screening model predicting cardiovascular invasion of early-and-intermediate hepatocyte cancer, Medicine 2018,97(49): 1-7). The scoring model is based on COX multifactor regression analysis, and Prothrombin Time (PT), aspartate aminotransferase level (AST) and BCLC (Barcelona liver cancer clinical staging system) B phase are selected as prediction factors to score patients. The sample size of the research population is too small, only 324 people are used for modeling queues, the representativeness is lacking, the characteristics of most of the population cannot be met, and the authenticity and universality of the research conclusion cannot be confirmed at present. Secondly, the research application is not intuitive enough, only research indexes are assigned, the model is not visualized, the result of the prediction model is not readable, and the clinical evaluation of patients is not convenient. Finally, the study population is limited, only the population receiving TACE combined RFA treatment in BCLC A-B stage is involved, and the application range is narrow.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention provides a system for predicting the vascular invasion probability of a primary liver cancer patient which cannot be resected by an operation.
Therefore, the invention adopts the following technical scheme:
a system for predicting the probability of vascular invasion in a non-surgically resectable primary liver cancer patient comprising:
a data collection module: data for obtaining the patient's etiology, BCLC staging, treatment modality, whether or not cirrhosis of the liver, tumor diameter, alpha-fetoprotein (AFP) levels, and C-reactive protein (CRP) levels;
a module for calculating the blood vessel invasion probability: assigning data of the etiology, BCLC stage, treatment mode, whether liver cirrhosis exists, tumor diameter, Alpha Fetoprotein (AFP) level and C-reactive protein (CRP) level by using an established visual Nomogram model for predicting the vascular invasion of a primary liver cancer patient which cannot be resected by an operation, calculating a total score, and obtaining corresponding vascular invasion probability according to the total score; the visualized Nomogram model is shown in FIG. 1; wherein HBV (hepatitis B) score is 46, non-HBV score is 0; cirrhosis was scored 93, non-cirrhosis was scored 0; BCLC stage B score of 41, BCLC 0-A stage score of 0, RFA treatment score of 80, TACE treatment score of 48, TACE combined with RFA treatment score of 0; the score of 62 is when the diameter of the tumor is more than or equal to 5cm, and the score of 0 is when the diameter of the tumor is less than 5 cm; the score of Alpha Fetoprotein (AFP) is more than or equal to 400ng/ml and is 52, and the score of Alpha Fetoprotein (AFP) is less than 400ng/ml and is 0; the score of C-reactive protein (CRP) is more than or equal to 5mg/L and is 100, and the score of C-reactive protein (CRP) is less than 5mg/L and is 0.
Preferably, the blood vessel invasion probability refers to the occurrence probability of blood vessel invasion within 1 year from the predicted date of the patient.
The invention also provides a method for predicting the blood vessel invasion probability of a primary liver cancer patient incapable of being resected by surgery, which is based on the system and comprises the following steps:
s-1. data acquisition step
Obtaining data on the patient's etiology, BCLC staging, treatment modality, whether cirrhosis of the liver, tumor diameter, alpha-fetoprotein (AFP) level, and C-reactive protein level;
s-2. data input step
Inputting the data collected in the step S-1 into the data collection module;
s-3, calculating the blood vessel invasion probability
And assigning the data of the etiology, the BCLC stage, the treatment mode, whether the liver is hardened, the tumor diameter, the Alpha Fetoprotein (AFP) level and the C-reactive protein (CRP) level by using the established visual Nomogram model for predicting the vascular invasion of the primary liver cancer patient which cannot be resected by surgery, calculating a total score, and obtaining the corresponding vascular invasion probability according to the total score.
Compared with a prediction model reported in the prior art (Yao Liu, et al. A new diagnosis model prediction macroscopic vascular invasion of early-intercipate hepatic vascular invasion. Medicine 2018,97(49):1-7), the model for predicting the vascular invasion probability of the primary liver cancer patient which cannot be resected by surgery has higher discrimination and more accurate result, thereby providing more valuable reference for clinical treatment means and medication selection.
Drawings
The invention is further described below with reference to the accompanying drawings.
FIG. 1 shows a visualized Nomogram model constructed by the present invention for predicting vascular invasion in patients with liver cancer.
Fig. 2 shows a visualized Nomogram model constructed by the present invention in a web page version.
FIG. 3 shows the ROC of the Nomogram model constructed in accordance with the present invention to predict vascular invasion in patients with liver cancer, where A is the curve of the modeling group and B is the curve of the validation group.
Fig. 4 shows tdROC of the Nomogram model constructed in the present invention for predicting vascular invasion in liver cancer patients, where a is the curve of the modeling group and B is the curve of the validation group.
FIG. 5 shows a calibration curve of a Nomogram model constructed in accordance with the present invention for predicting vascular invasion in patients with liver cancer, where A is the curve of the modeling group and B is the curve of the validation group.
Fig. 6 shows a clinical Decision Curve (DC) of the Nomogram model constructed in the present invention for predicting vascular invasion in patients with liver cancer, where a is a curve of the building block and B is a curve of the verification block.
FIG. 7 shows the Clinical Impact Curves (CIC) of the Nomogram model constructed in the present invention for predicting vascular invasion in patients with liver cancer, where A is the curve of the modeling group and B is the curve of the validation group.
FIG. 8 shows the K-M survival curves of the Nomogram model constructed in accordance with the present invention for predicting vascular invasion in patients with liver cancer, where A is the curve of the building block and B is the curve of the verification block. In the figure: the high risk is in the region numbered 1, the medium risk is in the region numbered 2 and the low risk is in the region numbered 3.
Fig. 9 shows MVI incidence curves for different treatment modalities, where a is the curve for the modeling group and B is the curve for the validation group.
Fig. 10 shows ROC curves for the constructed nomogrm model of the present invention and the model of Yao Liu of comparative example 1.
Detailed Description
The invention is illustrated below with reference to specific examples. It will be understood by those skilled in the art that these examples are for illustrative purposes only and are not intended to limit the scope of the present invention in any way.
The experimental methods in the following examples and examples are conventional unless otherwise specified. The reagent materials and the like used in the following examples are commercially available products unless otherwise specified.
Term abbreviations:
AFP (α -fetoprotein): alpha-fetoprotein;
alb (sodium album): a serum albumin;
alt (alanine aminotransferase): alanine aminotransferase;
ast (asparate aminotransferase): aspartate aminotransferase;
auc (area under curve): area under the curve;
BCLC (Barcelona clinical liver cancer): a Barcelona liver cancer clinical staging system;
CA-199(Carbohydrate antigen 199): carbohydrate antigen 199;
cea (carcinoembryonic antigen): carcinoembryonic antigen;
ci (confidence interval): a confidence interval;
cic (clinical impact curve): a clinical impact curve;
CRP (C-reactive protein) C-reactive protein;
ct (computed tomogry): computed tomography scanning;
dca (precision curve analysis): analyzing a decision curve;
HCC(hepatocellular carcinoma)：
HbeAg: hepatitis b e antigen;
HBsAg: hepatitis b surface antigen;
HBV: hepatitis B virus;
HBV-DNA: hepatitis b virus-deoxyribonucleic acid;
hr (hazard ratio): a risk ratio;
mri (magnetic resonance imaging): magnetic resonance imaging;
mvi (macroreticular vascular invasion): large vessel invasion;
NRI(net reclassification index)：
or (odds ratio): ratio of ratios;
plt (plateau): a platelet;
pta (prothrombin time activity): prothrombin activity;
rbc (red blood cell): red blood cells;
rfa (radio frequency offset): radiofrequency ablation;
roc (receiver operating curve): a receiver operating curve;
tace (transcatheter arterial chemobolization): transcatheter arterial chemoembolization;
tdROC (time-dependent receiver operating characteristics curve): a time dependent receiver operating curve;
wbc (white block cell): white blood cells.
Study example 1Construction of visual Nomogram model for predicting MVI of liver cancer patient
1 study object
Retrospectively collect patients who first confirmed primary liver cancer in Beijing Di Tan Hospital affiliated to capital medical university from 6 months in 2008 to 12 months in 2019, and divide the patients into a building group and a verification group according to a method of randomization grouping 7: 3.
2 diagnostic criteria
2.1 Primary liver cancer diagnostic criteria: a. pathological diagnosis criteria: the tissue specimen of the space occupying focus of the liver or the extrahepatic metastasis biopsy or the surgical excision is diagnosed as PLC through the pathological histology and/or cytology examination, and the PLC is the gold standard. b. Clinical diagnostic criteria: according to the primary liver cancer diagnosis and treatment standard formally revised by the ministry of health of the people's republic of China in 2019, when one item of the following conditions (a) or two items of the following conditions (b + are met), the clinical diagnosis of the primary liver cancer can be established:
typical primary liver cancer imaging characteristics:
a, if the occupying diameter of the liver is larger than 2cm, at least one imaging examination (dynamic enhanced MRI, enhanced CT, ultrasonic radiography and common display dynamic enhanced MRI) shows that the occupying area of the liver has typical liver cancer characteristics, and then hepatocellular carcinoma can be diagnosed;
b, if the occupied area of the liver is less than or equal to 2cm, at least two of the four imaging examinations are required to show that the occupied area of the liver has typical liver cancer characteristics, so that the primary liver cancer can be diagnosed to enhance the specificity of diagnosis.
If the above imaging can not exclude the diagnosis of liver cancer, 2-3 months of imaging follow-up or liver puncture biopsy can be performed.
② the AFP is larger than or equal to 400ug/L, namely the AFP is continuously over the normal value, at least one item of the above-mentioned image examination accords with the typical liver cancer characteristic, and can eliminate the AFP rise caused by other reasons, including pregnancy, germ system embryonic source tumor, active liver disease and upper digestive tract cancer, etc.
2.2 diagnostic criteria for vascular invasion: the international diagnostic criteria for vascular invasion are not well defined. According to the Chinese expert consensus on hepatocellular carcinoma amalgamation portal vein cancer embolus multidisciplinary diagnosis and treatment (2018 edition)[14]And Chinese specialist consensus on hepatocellular carcinoma with hepatic vein or inferior vena cava cancer embolus multidisciplinary diagnosis and treatment (2019 edition)[15]The CT or MRI is considered to show cancer emboli in the follow-up process and meet the imaging standard of the primary liver cancer: (1) the echogenicity of the embolus in the blood vessel is similar to that of the main tumor focus, and the embolus is mainly low echogenicity; (2) CT enhanced scanning shows that low-density filling defects can be seen in blood vessels; (3) MRI enhancement shows the filling defect in blood vessels, the performance is similar to CT, and the establishment of blood vessel invasion can be diagnosed by meeting 1 or more.
3 nano standard
3.1 inclusion criteria
(1) Those who meet the primary liver cancer diagnostic criteria;
(2) age 18-75 years, with unlimited genders;
(3) topical (TACE or RFA) treatment was performed.
3.2 exclusion criteria
(1) A patient with a vascular invasion has occurred;
(2) a patient with a distal metastasis;
(3) patients with metastatic liver cancer;
(4) patients with combined HIV and tuberculosis infection;
(5) severe diseases of the major organs such as the heart, brain, lung and kidney accompanied by dysfunction;
(6) patients with comorbid severe mental disorders;
(7) pregnant and lactating women;
(8) imperfect clinical data.
4 collecting and arranging data
The retrospective cohort study was performed with the time of local treatment at admission as the start of the study and 1 year follow-up observations, with the primary endpoint of the observation being the onset of vascular invasion and the secondary endpoint being death and distant metastasis. By consulting the electronic case record the following information:
(1) basic information and demographic characteristics: patient name, sex, age, date of first and last visit, family history, smoking history, drinking history, treatment method, medication history, etc.;
smoking history definition: smoking more than 10 cigarettes per day for more than one year;
definition of the history of drinking: drinking more than 20g per day, more than 5 times per week, for more than one year.
(2) Laboratory examination:
blood routine: white blood cells, red blood cells, platelets, etc.;
blood coagulation function: fibrinogen, etc.;
biochemical analysis: liver function, kidney function, etc.;
virology index: HBsAg, HBeAg, HBV-DNA;
tumor indexes are as follows: AFP, CEA, CA-199.
(3) Imaging data: b-ultrasound, CT, MRI, chest radiograph.
5 statistical methods
(1) For retrospective study, patients who met the criteria for enrollment were randomized into cohorts and validated groups for analysis based on follow-up, statistical analysis using R4.1.0, and rms, survivval, rmda, timeROC, etc. installation packages. P <0.05 is statistically significant;
(2) the counting data is represented by frequency, and the comparison is carried out by adopting' chi2Checking;
(3) screening out risk factors influencing vascular invasion of a liver cancer patient by single-factor and multi-factor analysis by using a Logistic regression model;
(4) drawing a visualized Nomogram model according to multiple factors by using an R language;
(5) a Kalplan-Merier method is adopted to draw a patient survival curve, and a Log-rank test method is adopted to compare survival rate;
(6) the discrimination of the model is checked by using a C index and the area under a time-dependent ROC curve, and the calibration degree of the model is checked by drawing a calibration curve;
(7) DCA and CIC were used to describe the clinical net benefit and performance improvement of this model.
6 results of the study
6.1 Baseline characteristic comparison of patients with liver cancer
1273 cases of primary liver cancer were included in the study, and all patients were divided into a building block and a validation block using random numbers according to the randomization principle, 893 cases (70%) and 380 cases (30%) of the validation block. The modeling group is used for fitting a Nomogram model for predicting the survival of the liver cancer patient, and the verification group is used for evaluating the prediction capability of the constructed model. Patients in the building and validation cohorts showed no difference in distribution in age, sex, personal history, oncology characteristics, and laboratory indices such as blood routine, liver function, and blood clotting function (P > 0.05). The clinical characteristics of the patients in the modeling and validation groups are shown in table 1.
TABLE 1 modeling and validation set of clinical characteristics
6.2 Logistic one-factor and multifactor regression analysis for influencing MVI incidence rate of liver cancer patients
As shown in table 2, the etiology (OR-1.799, 95% CI: 1.079-3.11), BCLC staging (OR-2.728, 95% CI: 1.344-5.626), cirrhosis (OR-3.276, 95% CI: 1.581-7.506), tumor size (OR-1.773, 95% CI: 1.08-2.883), AFP (OR-1.945, 95% CI: 1.278-2.946), CRP (OR-3.276, 95% CI: 2.256-4.763) were found to be independent risk factors for vascular invasion of liver cancer patients (P <0.05) by Logistic one-factor and multi-factor risk regression analysis; the mode of treatment (OR ═ 0.57, 95% CI: 0.391-0.832) is a protective factor. The OR values for vascular invasion in the RFA-treated group and TACE-RFA combination group were 1.645 (95% CI: 0.847-3.127; P ═ 0.133), 0.57 (95% CI: 0.391-0.832; P ═ 0.003), respectively, based on the TACE-treated group, i.e., the incidence of vascular invasion was reduced by 43% in the TACE-RFA combination group as compared to the TACE-treated group. These indices were also included and used to construct the Nomogram model.
TABLE 2 Logistic Single and Multi-factor regression analysis of independent risk factors affecting the incidence of 1-year vascular invasion in patients
6.3 construction of visual Nomogram model for predicting MVI of liver cancer patient
And importing the screened data of the relevant factors into open-source R software (3.6.3 version, https:// www.r-project. org) for data analysis to obtain the Etiology (ethology), whether cirrhosis (crihosis), BCLC (BCLC) staging, treatment mode (treatment), tumor diameter (tumor size), assigned values of AFP and CRP, and MVI incidence rate corresponding to the total score. Accordingly, a Nomogram is plotted, including the score, total score axis and MVI incidence axis for each relevant factor, as shown in fig. 1.
In FIG. 1, the uppermost horizontal axis is a one-factor score axis (0 to 100). The score axis, the total score axis (0-650) and the MVI occurrence rate axis of each single factor are sequentially arranged in parallel below the score axis. Wherein the causative HBV (hepatitis B) score is 46, the non-HBV score is 0; cirrhosis was scored 93, non-cirrhosis was scored 0; BCLC stage B score of 41, BCLC 0-A stage score of 0, RFA treatment score of 80, TACE treatment score of 48, TACE combined with RFA treatment score of 0; the score of 62 is when the diameter of the tumor is more than or equal to 5cm, and the score of 0 is when the diameter of the tumor is less than 5 cm; the AFP is more than or equal to 400ng/ml and the score is 52, and the AFP is less than 400ng/ml and is 0; the CRP is more than or equal to 5mg/L and the score is 100, and the CRP is less than 5mg/L and is 0. When the visual Nomogram constructed by the method is used, the score corresponding to each factor is found on the assigning axis, then the scores of all the factors are added, the corresponding point is found on the total score axis, a straight line perpendicular to the total score axis is drawn through the point and is intersected with the MVI occurrence rate axis below, and the numerical value of the intersection point is the probability of MVI occurrence.
The model constructed by the invention is already made into a webpage version for online use, and the MVI incidence rate of the patient is visually given according to the direct selection of options of the patient information (as shown in figure 2). The website is as follows: https:// hcnnomogran. shinyapps. io/DynNomo MVI/.
6.4 validation and evaluation of predictive models
6.4.1 discrimination evaluation of predictive models
In order to evaluate the discrimination of the visualized Nomogram model established by the invention, an ROC is drawn, as shown in FIG. 3, wherein A is the ROC of the modeling group, and B is the ROC of the verification group. The areas under the ROC curve of the model constructed by the invention for predicting 1-year vascular invasion of the liver cancer patient are respectively 0.785 and 0.762 in the building module and the verification group, and are both greater than 0.75, so that the method has better discrimination. For distinguishing whether MVI exists or not, the specificity and sensitivity of the building block are 79.1% and 65.1% respectively, and the specificity and sensitivity of the verification block are 77.6% and 66.2% respectively.
Considering the continuity of the survival time of the liver cancer, tdROC of the modeling group and the verification group are respectively drawn, and in the drawing, A is the tdROC of the modeling group, and B is the tdROC of the verification group. By analyzing the AUC of tdROC, the AUC value of the prediction model in the modeling group and the verification group at any survival time point of 4-10 months of liver cancer patients is higher than 0.7, which indicates that the model has good performance.
6.4.2 evaluation of the degree of calibration of the predictive model
The Calibration degree (Calibration) of the prediction model is an important index for evaluating the accuracy of the clinical prediction model. Calibration curves were plotted according to the Nomogram model, see fig. 5, where a is the calibration of the modeling group and B is the calibration curve of the validation group. FIG. 5 shows the degree of calibration between the actual incidence of 1 year vascular invasion and the predicted probability for liver cancer patients. The C index of the model group model is calculated to be 0.785, and the C index of the verification group model is 0.762 (see A and B of figure 5). Through drawing the calibration chart, the prediction model can be well fitted to the modeling group and the verification group easily, and the actual vascular invasion incidence rate of a patient can be well fitted.
6.4.3 clinical Decision Curve Analysis (DCA) of predictive models
The clinical decision curves for the modeling and validation sets are shown in FIG. 6, where A is the curve for the modeling set and B is the curve for the validation set. DCA is used to evaluate the net benefit under different clinical decisions with a certain threshold probability. Over DCA, it was shown that the use of the nomogr to predict MVI can increase more benefit than measures that treat all patients or none of them. The abscissa is the threshold probability: in the risk assessment tool, the probability that patient i is diagnosed with vascular invasion is denoted as Pi; when Pi reaches a certain threshold (denoted as Pt), it is defined as positive and therapeutic measures are taken. Some patients at this threshold will benefit from treatment and there will be non-patient treatment damage and untreated patient loss. While the ordinate is the net gain after the gain minus the loss. The black curved slope in a in fig. 6 represents the prediction model in the modeling block, and the bottom dark gray line represents the case when no treatment is given, with a net benefit of 0. The oblique light gray lines represent positive samples, all people receive treatment, the net benefit is the reverse oblique line with the negative slope, and through decision curve analysis, people can easily find that patients with the model prediction effective rate of 0.25-0.76 in the modeling group can all obtain the net benefit. Similar results were seen in the validation set, where the black curve in B of FIG. 6 represents the predictive model in the validation set, and patients with model prediction validation rates of 0.2-0.58 all had net benefit.
6.4.4 clinical impact evaluation of predictive models
The CIC of the visualized Nomogram model constructed by the method is shown in figure 7, wherein A is a curve of the modeling group, and B is a curve of the verification group. Predicting risk stratification of 1,000 persons using the prediction model, displaying "loss" and "benefit" axes, wherein the black solid line (high risk persons) in the graph represents the number of persons who are classified as positive (high risk) by the prediction model at each threshold probability for the building/verifying group; the gray dashed line is the true positive number for each threshold probability. CIC suggests that individualized models have a significant impact on clinical benefit.
6.5 Risk stratification analysis for predicting MVI in hepatocarcinoma patients by Nomogram model
The interquartile range scored according to the Nomogram model was divided into high, medium and low risk groups, with low risk group <139.3291 points; risk groups 139.3291-281.0578; high risk group > 281.0578 points. K-M survival curves are plotted, see FIG. 8, where A is the curve for the building set and B is the curve for the validation set. In the model group, the risk ratio (HR) values for medium-risk and high-risk overall survival were 2.93 (95% CI: 1.75-4.91; P <0.001) and 11.28 (95% CI: 6.87-18.52; P <0.001), respectively, with the low-risk group as a reference; in the validation group population, HR values for medium-risk and high-risk overall survival were 2.98 (95% CI: 1.24-7.17; P ═ 0.015) and 8.52 (95% CI: 3.62-20.09; P <0.001), respectively, with the low-risk group as a reference. It can be seen that the Nomogram model can clearly distinguish all patients according to different risk of death, whether it is a modeling group or a validation group.
6.6 MVI incidence analysis for different treatment modalities
The number of patients without MVI in 0-12 months of the modeling group and the verification group is shown in tables 3 and 4; the incidence of MVI in the two groups at 0-12 months is shown in FIG. 9.
TABLE 3 number of patients in the modeling group who did not develop MVI
Table 4 number of patients who did not develop MVI for the validation group of patients
In fig. 9, a shows that in the population of the building block, the incidence of vascular invasion was 34.6%, 27.5% and 17.1% for the TACE treatment modality, RFA treatment modality and TACE-RFA combination treatment modality at 12 months, respectively. The risk ratio (HR) values for RFA treatment and TACE-RFA combination treatment for vascular invasion were 1.37 (95% CI: 0.84-2.24; P ═ 0.206) and 0.60 (95% CI: 0.36-0.98; P ═ 0.038), respectively, with TACE treatment as reference. In fig. 9, B shows that the incidence of vascular invasion was 27.6%, 34.4% and 14.2% for TACE treatment modality, RFA treatment modality and TACE-RFA combination treatment modality at 12 months in the validation group population, respectively. For TACE treatment reference HR values for RFA treatment and TACE-RFA combination treatment vessel invasion were 0.79 (95% CI: 0.40-1.56; P ═ 0.506) and 0.38 (95% CI: 0.19-0.75; P ═ 0.004), respectively.
Comparative example 1Comparison of prediction results of a visual Nomogram model constructed by the invention and a model in the prior art
The visual Nomogram model constructed by the method is compared with a Yao Liu model (Yao Liu, et al. A new coding model prediction macro cellular estimation of early-interface geocellular medicine 2018,97(49):1-7) in a model prediction result scheme, and an ROC curve is drawn (see FIG. 10). Fig. 10 shows that the area under the model curve of the visualized Nomogram constructed by the invention is obviously higher than that of the model of Yao Liu, whether the model is a modeling group or a verification group, and the model is more differentiated.
Application example 1Clinical application of Nomogram model constructed by the invention
For example, one HBV-infected (46 min) liver cancer patient with cirrhosis-based (93 min) in BCLC stage B (41 min) received TACE in combination with RFA treatment (0 min) with a tumor diameter <5cm (0 min), AFP > 400ng/ml (52 min) and CRP > 5mg/L (100 min), a total score equal to 332, corresponding to a predicted incidence of MVI of 52%.
Claims (3)
1. A system for predicting the probability of vascular invasion in a non-surgically resectable primary liver cancer patient comprising:
a data collection module: data for obtaining the patient's etiology, BCLC staging, treatment modality, whether or not cirrhosis, tumor diameter, alpha-fetoprotein levels, and C-reactive protein levels;
a module for calculating the blood vessel invasion probability: assigning the data of the etiology, BCLC stage, treatment mode, whether liver cirrhosis exists, tumor diameter, alpha fetoprotein level and C-reactive protein level by using an established visual Nomogram model for predicting the vascular invasion of the primary liver cancer patient which cannot be resected, calculating a total score, and obtaining a corresponding vascular invasion probability according to the total score; the visualized Nomogram model is shown in FIG. 1; wherein, the HBV score is 46, and the non-HBV score is 0; cirrhosis was scored 93, non-cirrhosis was scored 0; BCLC stage B score of 41, BCLC 0-A stage score of 0, RFA treatment score of 80, TACE treatment score of 48, TACE combined with RFA treatment score of 0; the score of 62 is when the diameter of the tumor is more than or equal to 5cm, and the score of 0 is when the diameter of the tumor is less than 5 cm; the score of the alpha fetoprotein level is 52 when the alpha fetoprotein level is more than or equal to 400ng/ml, and the score of the alpha fetoprotein level is 0 when the alpha fetoprotein level is less than 400 ng/ml; the C-reactive protein level is greater than or equal to 5mg/L and the score is 100, and the C-reactive protein level is less than 5mg/L and is 0.
2. The system of claim 1, wherein the blood vessel invasion probability refers to the probability of occurrence of blood vessel invasion within 1 year from the predicted date of the patient.
3. A method of predicting the probability of vascular invasion in a non-surgically resectable primary liver cancer patient, said method being based on the system of claim 1 or 2, comprising the steps of:
s-1. data acquisition step
Obtaining data on the patient's etiology, BCLC staging, treatment modality, whether or not cirrhosis of the liver, tumor diameter, alpha-fetoprotein level, and C-reactive protein level;
s-2. data input step
Inputting the data collected in the step S-1 into the data collection module;
s-3, calculating the blood vessel invasion probability
And assigning the data of the etiology, the BCLC stage, the treatment mode, whether the liver is hardened, the tumor diameter, the alpha fetoprotein level and the C-reactive protein level by using the established visual Nomogram model for predicting the vascular invasion of the primary liver cancer patient which cannot be resected, calculating a total score, and obtaining the corresponding vascular invasion probability according to the total score.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111455870.4A CN114155956B (en) | 2021-12-02 | 2021-12-02 | System and method for predicting blood vessel invasion probability of primary liver cancer patient incapable of being resected by surgery |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111455870.4A CN114155956B (en) | 2021-12-02 | 2021-12-02 | System and method for predicting blood vessel invasion probability of primary liver cancer patient incapable of being resected by surgery |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114155956A true CN114155956A (en) | 2022-03-08 |
| CN114155956B CN114155956B (en) | 2022-09-20 |
Family
ID=80455710
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111455870.4A Active CN114155956B (en) | 2021-12-02 | 2021-12-02 | System and method for predicting blood vessel invasion probability of primary liver cancer patient incapable of being resected by surgery |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114155956B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115187512A (en) * | 2022-06-10 | 2022-10-14 | 珠海市人民医院 | Hepatocellular carcinoma great vessel invasion risk prediction method, system, device and medium |
| CN115985497A (en) * | 2022-12-16 | 2023-04-18 | 首都医科大学附属北京地坛医院 | System for predicting prognosis of non-operation treatment primary liver cancer patient mainly based on platelet/spleen major-diameter ratio |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160258958A1 (en) * | 2014-12-08 | 2016-09-08 | Berg Llc | Use of markers including filamin a in the diagnosis and treatment of prostate cancer |
| CN107180154A (en) * | 2017-02-08 | 2017-09-19 | 冯德昭 | The method and system of patients with orthotopic liver transplantation prognosis situation packet is carried out to the patients with hepatocellular carcinoma of single tumour |
| CN108565024A (en) * | 2018-03-19 | 2018-09-21 | 首都医科大学附属北京地坛医院 | The system of risk of recurrence in a kind of postoperative 1 year of the HBV correlation primary small hepatocellular carcinomas of determining single-shot |
| CN110577998A (en) * | 2019-01-31 | 2019-12-17 | 上海交通大学医学院附属仁济医院 | Construction of molecular model for predicting postoperative early recurrence risk of liver cancer and application evaluation thereof |
| CN110660481A (en) * | 2019-09-27 | 2020-01-07 | 颐保医疗科技(上海)有限公司 | Artificial intelligence technology-based primary liver cancer recurrence prediction method |
| CN111584082A (en) * | 2020-05-15 | 2020-08-25 | 高春芳 | Establishment and application of primary hepatocellular carcinoma microvascular invasion regression prediction model based on clinical examination multidimensional data |
| CN112017791A (en) * | 2020-04-24 | 2020-12-01 | 首都医科大学附属北京地坛医院 | System for determining prognosis condition of liver cancer patient based on artificial neural network model |
| CN112289450A (en) * | 2020-12-25 | 2021-01-29 | 江苏高美基因科技有限公司 | Prediction system for prognosis survival period of intrahepatic cholangiocellular carcinoma patient |
| CN113362885A (en) * | 2021-06-03 | 2021-09-07 | 复旦大学附属中山医院 | Method for establishing prognosis model of HAIC (liver cancer-associated syndrome) treatment advanced liver cancer patient |
-
2021
- 2021-12-02 CN CN202111455870.4A patent/CN114155956B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160258958A1 (en) * | 2014-12-08 | 2016-09-08 | Berg Llc | Use of markers including filamin a in the diagnosis and treatment of prostate cancer |
| CN107180154A (en) * | 2017-02-08 | 2017-09-19 | 冯德昭 | The method and system of patients with orthotopic liver transplantation prognosis situation packet is carried out to the patients with hepatocellular carcinoma of single tumour |
| CN108565024A (en) * | 2018-03-19 | 2018-09-21 | 首都医科大学附属北京地坛医院 | The system of risk of recurrence in a kind of postoperative 1 year of the HBV correlation primary small hepatocellular carcinomas of determining single-shot |
| CN110577998A (en) * | 2019-01-31 | 2019-12-17 | 上海交通大学医学院附属仁济医院 | Construction of molecular model for predicting postoperative early recurrence risk of liver cancer and application evaluation thereof |
| CN110660481A (en) * | 2019-09-27 | 2020-01-07 | 颐保医疗科技(上海)有限公司 | Artificial intelligence technology-based primary liver cancer recurrence prediction method |
| CN112017791A (en) * | 2020-04-24 | 2020-12-01 | 首都医科大学附属北京地坛医院 | System for determining prognosis condition of liver cancer patient based on artificial neural network model |
| CN111584082A (en) * | 2020-05-15 | 2020-08-25 | 高春芳 | Establishment and application of primary hepatocellular carcinoma microvascular invasion regression prediction model based on clinical examination multidimensional data |
| CN112289450A (en) * | 2020-12-25 | 2021-01-29 | 江苏高美基因科技有限公司 | Prediction system for prognosis survival period of intrahepatic cholangiocellular carcinoma patient |
| CN113362885A (en) * | 2021-06-03 | 2021-09-07 | 复旦大学附属中山医院 | Method for establishing prognosis model of HAIC (liver cancer-associated syndrome) treatment advanced liver cancer patient |
Non-Patent Citations (10)
| Title |
|---|
| DONG DONG ZHOU；XIAO LI LIU；XIN HUI WANG: "A prognostic nomogram in AFP-negative hepatocellular carcinoma based on LASSO Cox regression", 《PRIMARY RESEARCH》 * |
| JIANG JIANG , HAO HU, RONG LIU: "Nomogram for individualized prediction of recurrence after postoperative adjuvant TACE for hepatitis B virus-related hepatocellular carcinoma", 《MEDICINE》 * |
| YIFEI WANG等: "Prognostic nomograms on inflammatory markers in hepatitis B virus-related hepatocellular carcinoma patients who underwent hepatectomy", 《CANCER BIOLOGY & MEDICINE》 * |
| 何羽童等: "基于临床实验室指标的HBV相关性原发性肝癌微血管侵犯预测模型的建立", 《临床检验杂志》 * |
| 廖诗琪等: "乙型肝炎病毒相关性肝细胞肝癌微血管浸润的术前临床与CT影像特征分析及Nomogram预测模型的建立", 《临床放射学杂志》 * |
| 李金鹏等: "NLR对肝癌TACE患者预后影响的分析", 《中华肿瘤防治杂志》 * |
| 王维；金宗睿；吴国林；王珏: "预测大肝癌患者预后的列线图：一项基于SEER数据库的研究", 《中国普外基础与临床杂志》 * |
| 胡月雷等: "单发肝细胞癌发生微血管侵犯的影响因素及预后分析", 《临床肝胆病杂志》 * |
| 谭君心等: "血小板淋巴细胞比值和肿瘤直径与肝细胞癌微血管侵犯的相关性", 《标记免疫分析与临床》 * |
| 马依迪丽•尼加提 迪丽阿热姆•艾海提: "诺模图(nomogram)预测浸润性乳腺癌脉管侵犯的初步研究", 《影像诊断与介入放射学》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115187512A (en) * | 2022-06-10 | 2022-10-14 | 珠海市人民医院 | Hepatocellular carcinoma great vessel invasion risk prediction method, system, device and medium |
| CN115187512B (en) * | 2022-06-10 | 2024-01-30 | 珠海市人民医院 | Method, system, device and medium for predicting invasion risk of large blood vessel of hepatocellular carcinoma |
| CN115985497A (en) * | 2022-12-16 | 2023-04-18 | 首都医科大学附属北京地坛医院 | System for predicting prognosis of non-operation treatment primary liver cancer patient mainly based on platelet/spleen major-diameter ratio |
| CN115985497B (en) * | 2022-12-16 | 2023-08-08 | 首都医科大学附属北京地坛医院 | System for predicting prognosis of patient with non-operative treatment primary liver cancer based on platelet/spleen aspect ratio value |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114155956B (en) | 2022-09-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wong et al. | Elevated alpha-fetoprotein: differential diagnosis-hepatocellular carcinoma and other disorders | |
| Huang et al. | Value of free/total prostate-specific antigen (f/t PSA) ratios for prostate cancer detection in patients with total serum prostate-specific antigen between 4 and 10 ng/mL: A meta-analysis | |
| Carlsson et al. | Circulating tumor microemboli diagnostics for patients with non–small-cell lung cancer | |
| Xu et al. | Serum Golgi protein 73 (GP73) is a diagnostic and prognostic marker of chronic HBV liver disease | |
| Swensen et al. | CT screening for lung cancer: five-year prospective experience | |
| CN114155956B (en) | System and method for predicting blood vessel invasion probability of primary liver cancer patient incapable of being resected by surgery | |
| Ye et al. | Circulating tumor cells as a potential biomarker for postoperative clinical outcome in HBV-related hepatocellular carcinoma | |
| Sheron et al. | Developing a ‘traffic light’test with potential for rational early diagnosis of liver fibrosis and cirrhosis in the community | |
| Xing et al. | Clinical performance of α-L-fucosidase for early detection of hepatocellular carcinoma | |
| Yuyun et al. | Predictive value of the red blood cell distribution width-to-platelet ratio for hepatic fibrosis | |
| Hogea et al. | Bronchoalveolar lavage: role in the evaluation of pulmonary interstitial disease | |
| Chrostek et al. | Noninvasive indirect markers of liver fibrosis in alcoholics | |
| CN114141371B (en) | System for predicting prognosis of liver cancer associated with tumor-related anemia patient caused by hepatitis B virus infection | |
| Crane et al. | Retrospective validation of the pulmonary embolism rule-out criteria rule in ‘PE unlikely’patients with suspected pulmonary embolism | |
| Peng et al. | Spleen stiffness and volume help to predict posthepatectomy liver failure in patients with hepatocellular carcinoma | |
| Li et al. | Prognostic nomogram for overall survival in extranodal natural killer/T-cell lymphoma patients | |
| Wang et al. | The pretreatment platelet count is an independent predictor of tumor progression in patients undergoing transcatheter arterial chemoembolization with hepatitis B virus-related hepatocellular carcinoma | |
| Zhang et al. | Diagnostic accuracy and prognostic significance of osteopontin in liver cirrhosis and hepatocellular carcinoma: a Meta-analysis | |
| Maxwell et al. | CT-guided transthoracic needle aspiration biopsy of subsolid lung lesions | |
| Bittla et al. | Exploring Circulating Tumor DNA (CtDNA) and Its Role in Early Detection of Cancer: A Systematic Review | |
| CN115985497B (en) | System for predicting prognosis of patient with non-operative treatment primary liver cancer based on platelet/spleen aspect ratio value | |
| Ren et al. | Evaluation of the Alpha-Fetoprotein Model for Predicting Recurrence and Survival in Patients with Hepatitis B Virus (HBV)–Related Cirrhosis Who Received Liver Transplantation for Hepatocellular Carcinoma | |
| Wang et al. | New insights into a microvascular invasion prediction model in hepatocellular carcinoma: A retrospective study from the SEER database and China | |
| Lee et al. | Complete blood count reflects the degree of oesophageal varices and liver fibrosis in virus‐related chronic liver disease patients | |
| Li et al. | Transitional cell carcinoma associated with aristolochic acid nephropathy: most common cancer in chronic hemodialysis patients in China |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |