CN105319364A - Combined diagnosis marker for predicting small hepatocellular carcinoma relapse - Google Patents
Combined diagnosis marker for predicting small hepatocellular carcinoma relapse Download PDFInfo
- Publication number
- CN105319364A CN105319364A CN201510744435.1A CN201510744435A CN105319364A CN 105319364 A CN105319364 A CN 105319364A CN 201510744435 A CN201510744435 A CN 201510744435A CN 105319364 A CN105319364 A CN 105319364A
- Authority
- CN
- China
- Prior art keywords
- liver cancer
- gene
- hepatocellular carcinoma
- relapse
- recurrence
- 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
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/48—Biological material, e.g. blood, urine; Haemocytometers
- G01N33/50—Chemical analysis of biological material, e.g. blood, urine; Testing involving biospecific ligand binding methods; Immunological testing
- G01N33/53—Immunoassay; Biospecific binding assay; Materials therefor
- G01N33/574—Immunoassay; Biospecific binding assay; Materials therefor for cancer
- G01N33/57407—Specifically defined cancers
- G01N33/57438—Specifically defined cancers of liver, pancreas or kidney
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/48—Biological material, e.g. blood, urine; Haemocytometers
- G01N33/50—Chemical analysis of biological material, e.g. blood, urine; Testing involving biospecific ligand binding methods; Immunological testing
- G01N33/53—Immunoassay; Biospecific binding assay; Materials therefor
- G01N33/574—Immunoassay; Biospecific binding assay; Materials therefor for cancer
- G01N33/57484—Immunoassay; Biospecific binding assay; Materials therefor for cancer involving compounds serving as markers for tumor, cancer, neoplasia, e.g. cellular determinants, receptors, heat shock/stress proteins, A-protein, oligosaccharides, metabolites
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Immunology (AREA)
- Engineering & Computer Science (AREA)
- Urology & Nephrology (AREA)
- Chemical & Material Sciences (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- Hematology (AREA)
- Cell Biology (AREA)
- Microbiology (AREA)
- General Health & Medical Sciences (AREA)
- Oncology (AREA)
- Hospice & Palliative Care (AREA)
- Pathology (AREA)
- Food Science & Technology (AREA)
- Medicinal Chemistry (AREA)
- Physics & Mathematics (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Biotechnology (AREA)
- General Physics & Mathematics (AREA)
- Gastroenterology & Hepatology (AREA)
- Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The invention discloses a combined diagnosis marker for predicting small hepatocellular carcinoma relapse. Based on an LASSO Cox regression model, the inventor confirms and establishes a small hepatocellular carcinoma relapse prediction model based on 6 gene signatures (including CD147, IL-7, Ki67, MIF, P97 and PD-L1) through analyzing a large number of cases; the risk ratio, defined by a 6 gene signature classifier, of a high relapse risk crowd to a low relapse risk crowd is 2.721 (95% CI: 1.633-4.532; P is smaller than 0.0001); the analysis of an internal validation group and an external validation group has the same result. The experiment proves that the classifier based on the 6 gene signatures can effectively predict different relapse risks of patients suffering from small hepatocellular carcinoma, the traditional diagnosis accuracy rate, predicted through clinicopathological parameters, of small hepatocellular carcinoma relapse is increased, the relapse risk of AFP-negative patients suffering from small hepatocellular carcinoma can be well judged, and the combined diagnosis marker has potential clinical value in the individualized treatment of small hepatocellular carcinoma.
Description
Technical field
The present invention relates to one group of diagnosing tumor mark, particularly for predicting the Combining diagnosis mark that small liver cancer recurs.
Background technology
Liver cancer is one of common malignant tumour of digestive system, and the incidence of disease occupies the 6th, the world, and mortality ratio occupies the 3rd, about has 250,000 people to die from liver cancer every year.Liver cancer is hidden due to onset, and grade malignancy is high, and cure rate is low, is one of malignant tumour of serious threat human health and quality of life, and the incidence of disease of liver cancer is all rising every year constantly.China is the country that liver cancer annual new cases in the whole world is maximum, accounts for the half of whole world total incidence, though and American-European countries's incidence of disease is not high is increasing year by year yet.Hepatocellular carcinoma (heptocellularcarcinoma, HCC) is the topmost histological type of liver cancer, accounts for the 70%-85% of Primary Hepatic malignant tumour, has accounted for about 90% especially in China.Up to now, the pathogenesis of HCC is still indefinite, but the absorption of virus infections, aflatoxin B1 and excessive drinking etc. are considered to the main paathogenic factor of HCC.Wherein 10%-40% chronic hepatitis b virus carrier develops into hepatocellular carcinoma the most at last, and the whole world is estimated to exceed 100 ten thousand patients every year and dies from hepatitis B virus (hepatitisBvirus, the HBV) liver cancer relevant with hepatitis C virus.At China and Japan, 75%-80%HCC morbidity is in close relations with liver chronic viral infection, and wherein hepatitis B virus infection rate accounts for 50%-55%, and in world wide, the HCC case of 80%-90% is associated with cirrhosis.Therefore, liver HBV infection and cirrhosis are the topmost high risk factor of HCC and precancerous lesion process.
In recent years, along with the development of Protocols in Molecular Biology and Imaging Technology, very much progress is achieved to the early diagnosis of liver cancer and treatment, makes the clinical diagnosis of liver cancer and result for the treatment of all obtain remarkable improvement.According to the standard that Chinese Medical Association's surgery hepatic surgery group calendar year 2001 formulates, the liver cancer of single-shot diameter≤2cm is decided to be Small Liver Cancer, the diameter > 2cm of single-shot and the liver cancer of≤5cm is small liver cancer.And clinicopathologia to define small liver cancer be single or two cancerous node diameter≤3cm.Clinical pathological study data confirms: the small liver cancer prognosis of more than 3cm is relatively poor, portal vein tumor thrombus and stellate ganglion incidence all higher, then histological is better for the small liver cancer of≤3cm, tumour majority has coating, DNA content is mostly dliploid, rare cancer embolus and satellite stove, these prompting tumours are the main period that biological characteristics occurs obviously to change at below 3cm.From the 70-80 age in 20th century, along with the widespread use of serum alpha-fetoprotein (AFP), real-time ultrasound video picture and CT, substantially increase the early diagnostic rate of liver cancer, a lot of small liver cancer case is found clinical.Small liver cancer belongs to early liver cancer more, is the focus of liver cancer clinical research always.The treatment of small liver cancer mainly emphasizes radical-ability, and excision is considered to " goldstandard " of small liver cancer radical treatment, is the first-selected treatment means of small liver cancer.But small liver cancer postoperative 5 years survival rates still rest between 50%-60%, and do not have great breakthrough all the time, postoperative recurrence and transfer remain the main cause of small liver cancer prognosis mala.
The source one of liver cancer recurrence is that multicenter occurs, namely after liver cancer radical excision, due to the soil (cirrhosis) of liver cancer growth and existing of other carcinogenic factor, then the tumour sending out new; Two is that single centre occurs, and namely in preoperative, the art of original removal of lesions, cancer cell trans-portal vein approach is sent out, and recurrence and extrahepatic metastases in liver occur.At present on the research affecting liver cancer survival analysis, mainly concentrate on the analysis to clinical pathological factors such as tumor size, number, differentiation degree, vascular invasion, Level of Alpha Fetoproteins, and these still can not be predicted more accurately to the recurrence after liver cancer excision.Therefore, for this sciences problems of liver cancer recurrence, how after Hemi-Resection, tumor recurrence is effectively predicted, the postoperative tumor recurrence that how to prevent liver cancer all has great clinical meaning to improving the curative effect of liver cancer, explore and set up simple, fast, the molecular prediction index of the high and high specificity of susceptibility and new therapy target become improve clinical efficacy in the urgent need to.
The recurrence of liver cancer is the complex process of a multi-step, multifactor participation, relate to the oncobiology various aspects such as tumor cell adhesion, extracellular matrix degradation, cell migration, cell proliferation, Tumor angiogenesis, can not separately only from analyzing liver cancer recurrence on the one hand.The recurrence of liver cancer is relevant with clinicopathologic stage, tumor capsule, tumor size, number, vessel invasion and degree of cirrhosis etc., these results are just based upon liver cancer clinical pathology and cellular level is observed and on the basis of research, be difficult to the prediction its tendency of recurrence being made in early days to individuation before small liver cancer (especially unijunction joint) recurrence.Research in recent years finds to affect liver cancer recurrence metastatic gene, albumen plays an important role in its biological behaviour, but current great majority research is single factor test research mode, concentrate research one or a few gene, the expression of albumen, only only with one or a few gene, the expression of albumen definitely cannot reflect the molecular biological characteristics of hepatoma Metastasis, therefore the method for multifactor associating should be adopted further comprehensively to analyze multiple index that tumor invasion and metabasis is correlated with and the relation that small liver cancer recurs, distinguish the inner link between each index of correlation, thus set up a suitable pattern to predict the recurrence after small liver cancer operation in patients.
Recently, part research finds that gene expression label, epigenetic molecular labeling and microRNA can the recurrence of predicting liver cancer patient and prognosis very well.But, these based on gene label detection due to lab platform, flesh tissue obtain, repeatability, the reason such as testing cost and method popularization make it be difficult to widespread use clinically.And method for immunohistochemical detection because its operation is simple and easy, application is wide, expense is low, qualitative highly sensitive, location directly the factor such as accurate obtain the favor of researcher, be used to the molecule parting of various tumour, comprise hepatocellular carcinoma.These above researchs are all the results obtained based on large small liver cancer case, and the biological characteristics of small liver cancer and Large HCC also exist certain difference, small liver cancer (especially the single-shot diameter≤3cm of the HBV positive) this special subgroup is not but also predicted very well to the relevant report of Patients on Recurrence.
Summary of the invention
The object of the invention is to mention a kind of for predicting the Combining diagnosis mark that small liver cancer recurs, the recurrence of this mark to small liver cancer has good predictive value.
Inventor is by analyzing a large amount of case, and in training group case, LASSOCox regression model is established and constructed the small liver cancer recurrence prediction model based on 6 gene labels (comprising CD147, IL-7, Ki67, MIF, P97 and PD-L1); The high risk of recurrence crowd of 6 gene label sorter definition is 2.721 (95%CI:1.633-4.532 with the danger ratio of low risk of recurrence crowd; P<0.0001); Internal verification row (167 example) of forming a team are carried out to same analysis and also obtain similar result.In order to verify whether 6 gene label sorters have same prediction efficiency at different crowd, this sorter is applied to and forms a team in row from the not concentric separate outer checking including 347 routine small liver cancers by inventor.Find:
1) the high risk of recurrence crowd of sorter definition is 1.868 (95%CI:1.334-2.617 with the danger ratio of low risk of recurrence crowd; P<0.0001);
2) by getting rid of the synergy of clinicopathologic features, it is the powerful independent prognostic factor affecting small liver cancer recurrence that multivariate survival analysis result shows 6 gene label sorters, no matter be in training group and internal verification group, or all obtain similar result (P<0.0001) in external certificate group;
3) layering survival analysis result shows the prognostic model that 6 gene label sorters remain small liver cancer clinically and statistically significance; ROC tracing analysis shows 6 gene label sorters and have better accuracy (AUC:0.686vs0.606) than clinical clinical risk factors in the recurrence of prediction small liver cancer;
4) prognostic value (AUC=0.703) that 6 gene label sorters and clinical pathology high risk factor can improve parameter is combined; By constructing the alignment diagram that two integrate 6 gene label sorters and clinical high risk factor: AFP negative group and the positive group of AFP; The nomographic prediction effect that rule figure shows AFP negative group is better; Time dependent ROC tracing analysis display column line chart has higher prediction accuracy the patient of AFP negative.
Accompanying drawing explanation
Fig. 1 is that SABC detects the expression of 30 gene labels at small liver cancer tissue;
Fig. 2 is the cut off value that X-tile analysis software determines 30 gene label Immunohistochemical Expressions; Case is divided into two subgroups to carry out log-rank inspection by X-tile software, and the chi-square value that its inspection produces is as shown in the left side; All cases are divided into blue and grey two groups (Y-axis represents case load for middle square chart, the expression mark of X-axis expressive notation thing) according to the expression of gene label by X-tile software; The right display Kaplan-Meier survivorship curve figure;
Fig. 3 builds the small liver cancer prognosis classification device based on 6 biomarkers; (A) the biomarker LASSO coefficient distribution overview that 30 liver cancer is relevant: what perpendicular line showed is the threshold value determined by 10 times of cross validations; (B) the Ten-time cross validation of the tuning Selecting parameter of LASSO model: solid vertical line represents partial likelihood deviation ± SE (standard error), the optimum value that vertical dotted line representative is obtained by minimum sandards and 1-SE standard;
Fig. 4 is the risk score based on 6 gene label sorters of training group, internal verification group and external certificate group, time dependent ROC curve and Kaplan-Meier survivorship curve;
Fig. 5 is the Kaplan-Meier survival analysis of the clinicopathologic features layering according to the routine small liver cancer patient of 6 gene label detection of classifier 682;
Fig. 6 is the Kaplan-Meier survival analysis of the clinicopathologic features layering according to the routine small liver cancer patient of 6 gene label detection of classifier 682;
Fig. 7 is the prediction accuracy that ROC curve compares 6 gene label sorters, clinicopathologic features and individual gene label;
Fig. 8 is the tumor recurrence risk of alignment diagram prediction AFP negative and AFP sun two groups of small liver cancer patients.
Embodiment
Below in conjunction with experiment, further illustrate technical scheme of the present invention.
Sample data
Collect non-metastatic small liver cancer case 682 example of continuous print definitive pathological diagnosis and excision.Definitive pathological diagnosis small liver cancer 335 example during wherein coming from pathology department of Tumor Hospital Attached to Zhongshan Univ. year Dec in Dec, 1998 to 2008.The small liver cancer case of 335 routine Tumor Hospital Attached to Zhongshan Univ. is divided into by allocation order at random that generate by computer: Training Team 168 example, internal verification group 167 example.After internal verification, small liver cancer 347 example that during have collected again year September in July, 2000 to 2009, other hospitals make a definite diagnosis is as a separate outer checking group.The small liver cancer of primary excision (non-radio frequencies melts or liver transfer operation) case is only had just to include this research in.Inclusive criteria comprises: tumour single-shot, diameter of tumor≤3cm, and serum HBV is positive, and treatment is first excision, not transfer and residual tumor, and preoperative non-row supplemental treatment, has complete Follow-up Data.All case samples are all fixed through 10% formalin, paraffin embedding, and section is made conventional H & E and dyeed.All cases all re-start Histopathology assessment, and histological grade is with reference to Edmonson and Steiner standard.
Follow up a case by regular visits to
After excision, patient follows up a case by regular visits in outpatient service routinely, follows up a case by regular visits to and is spaced apart the 2-6 month, and the project of following up a case by regular visits to comprises Serum AFP, Abdominal B type ultrasonography and chest x-ray sheet.Process of following up a case by regular visits to when suspecting tumor recurrence then further row CT and MRI check, definitive pathological diagnosis is carried out in row aspiration biopsy if desired.Date of surgery is defined as to the tumour first time interval of examining the recurrence date life cycle without recurrence.TS Overall survival is defined as date of surgery and dies from liver cancer to patient or finally follow up a case by regular visits to the time interval on date.
H & E dyes
Morphological observation evaluates whether tumor grade, tumor capsule, tumour be downright bad, tumor vessel Infiltrating and organization chip representative region location to adopt conventional H & E colouring method to carry out.
Organization chip makes
(1) " acceptor " wax stone is made
The paraffin mould of come card LEICAPARAPLAST*HM is made " acceptor " wax stone that size is about 4.0cm × 2.5cm × 1.5cm, and with repairing wax machine by for subsequent use for wax stone surface equating.
(2) " donor " tissue is selected
Select pathology typical parts under microscope and carry out mark on slide, then section and Paraffin tissue block being mated, on Paraffin tissue block, making corresponding mark according to slide lesions position.
(3) organization chip square formation is made
" acceptor " wax stone made is fixed in the wax stone groove on organization chip maker, burrow at " acceptor " wax stone with the thin piercing needle on organization chip machine, from " donor " wax stone, take out required tissue with thick needle, insert in the hole of " acceptor " wax stone.Repeat above step, one by one selected wax stone tissue is inserted in " acceptor " wax stone, be arranged in micro-assembly robot array, namely make chip wax stone.(4) chip block sections histotome carries out serial section to organization chip wax stone, thickness is about 4um, the section made is made H & E to dye, whether the microarray made by inspection is consistent with the tissue line in designed array and whether contain tumor tissues.
Immunohistochemical staining
EnVision two step method is adopted to carry out immunohistochemical staining:
1) 3-5 μm of paraffin section 56 DEG C of incubators are baked sheet and are spent the night;
2) dimethylbenzene dewaxing and graded ethanol aquation;
3) distilled water rinsing 1 minute;
4) 0.3%H
2o
2soaking at room temperature 10 minutes, to block endogenous peroxydase;
5) distilled water rinsing 1 minute;
6) antigen retrieval:
The reparation of all antigen all adopts Pressure method method: appropriate antigen retrieval buffers (working fluid) (the reparation liquid that often kind of antibody uses refers to table 4) is injected Stainless steel pressure cooking-vessel and is heated to boiling.Section be positioned over and repair in liquid, add after cap gives vent to anger until pressure cooker, low fire maintains 2:30 minute, by pressure cooker end from thermal source, is cooled to room temperature;
7) pH7.2-7.4PBS damping fluid rinsing 1 minute;
8) with sheep blood serum hatch 30 minutes non-specific painted to remove;
9) drip primary antibodie, incubated at room is spent the night;
10) pH7.2-7.4PBS damping fluid rinsing (2 minutes × 5);
11) drip EnVision bis-to resist, incubated at room 60 minutes;
12) pH7.2-7.4PBS damping fluid rinsing (2 minutes × 5);
13) DAB colour developing 2-10 minute, clear water rinses cessation reaction;
14) haematoxylin redyes 1-2 minute, and hydrochloride alcohol breaks up 1-3 second;
15) fully rinse in flowing water;
16) sheet baked by 37 DEG C of incubators, and dimethylbenzene is transparent, neutral gum mounting.
ImmunohistochemistryResults Results is evaluated
The scoring of 30 gene labels uses semi-quantitative method: the ratio that calculating positive tumor cell accounts for total tumour cell is multiplied by its staining power.Number percent shared by tumour cell dyeing is divided into semiquantitative Pyatyi hierarchy system: 0 is divided into all tumour cells to be all unstained; 1 is divided into the tumour cell of 1% to 10% to be colored; 2 are divided into the tumour cell of 11% to 25% to be colored; 3 are divided into the tumour cell of 26% to 50% to be colored; 4 are divided into the tumour cell of 51% to 75% to be colored; 5 are divided into the tumour cell of more than 76% to be colored.Staining power uses semiquantitative 4 grades of points-scoring systems: 0 is not dyeing; 1 is weak dyeing; 2 is moderate dyeing; 3 is strong dyeing.By two, ImmunohistochemistryResults Results is not known that the pathologist of patients clinical pathological data is marked and is produced.When the result of two pathologist is consistent, the net result of scoring is jointly discussed by the 3rd pathologist and rear decision is discussed.
The selection of SABC cut off value
The cut off value that SABC scoring is best is determined by X-tile analysis software (3.6.1 version, Yale University School of Medicine), and X-tile analysis software calculates based on the relation between the expression mark of each gene label and patient's DFS.X-tile software provides a relation more intuitively between method assessment variable and survival of patients.X-tile software can calculate the cut off value of its maximum chi-square value (i.e. minimum p value) and then automatic choice variable optimum according to Kaplan-Meier survival analysis and Log-rank inspection.
LASSO regretional analysis
LASSO (LeastAbsoluteShrinkageandSelectionOperator) is the homing method of a kind of popular higher-dimension prediction.It obtains the model of a comparatively refining by constructing a punishment number, it shrinks some coefficients simultaneously, and to set some coefficients be zero.LASSO is a kind of method that process has the Biased estimator of multi-collinearity data, and by the absolute value of regression coefficient, minimum and residual error two is taken advantage of with minimum, obtains the homing method of a compromise ridge regression and sub-set selection.This method is expanded and is widely used in the Cox proportional hazards regression models of high dimensional data survival analysis.In training group case, inventor applies LASSOCox regression model and chooses and contact tight and useful prognostic marker with small liver cancer, and constructs model (R software, the version 3 .0.1 of the prediction small liver cancer recurrence based on a polygenes label; " glmnet " software package).
Statistical analysis
Relatively two groups of data difference continuous variables use t inspection and classified variable to adopt Chi-square Test.Inventor uses Kaplan-Meier methods analyst variable and small liver cancer without recurring the correlativity between surviving, application log-rank check analysis survivorship curve.The accuracy of classifier based on the ROC tracing analysis clinicopathologic features that inventor's Applicative time relies on and many biomarkers.Inventor uses Cox regression model to carry out multivariate survival analysis, and Cox regression coefficient makes alignment diagram.Nomographic characteristic performance inquired into by application rule.Time Dependent ROC analyzes and is completed by R software " survivalROC " packet, and alignment diagram and rule utilize R software rms packet to complete, and other statistics is then completed by 3.0.1 version R software.Think there is statistical significance when bilateral P value is less than 0.05.
3.1 patients clinical pathological characters
The small liver cancer case of inventor to 682 routine excisions, definitive pathological diagnosis carries out allelic expression detection.All patients are the patient of long-term hepatitis b virus carrier and first operation Resection in Treatment.Wherein after 275 routine corrective surgeries through going through tumor recurrence, meta recurrence time is 38.5 months.Current research 545 liver cancer sample has been purchased.Table 2.2 summarizes the clinicopathological characteristics of three groups of Small Hepatocellular Carcinoma patients in this research.The clinicopathological characteristics of three groups of cases and RFS lead does not have notable difference (table 2.2).The modal cause of death of patient is tumor recurrence, transfer or serious cirrhosis.Make a definite diagnosis the follow-up of patients with recurrent to control
The clinical pathologic characteristic of table 2.2 small liver cancer patient
* P value is obtained by the difference between comparative training group and checking group.
3.2 candidate gene labels are at the expression of small liver cancer tissue
According to bibliographical information in the past, inventor chooses 30 gene labels relevant to liver cancer genesis and development and carries out SABC detection.30 gene labels at the typical expression of Tissues of Hepatocellular Carcinoma IHC as shown in Figure 1.According to the research before inventor, in order to allow the predictive value of each label maximize, inventor applies the cut off value that each gene label in training group determined by X-tile software.X-tile software by the relation determination variable of situational variables and survival of patients single, cut off value intuitively, X-tile software can according to the cut off value of maximum chi-square value and the minimum automatic choice variable of P value.X-tile analysis software determines the cut off value (Fig. 2) of 30 gene label Immunohistochemical Expressions: case is divided into two subgroups to carry out log-rank inspection by X-tile software, and the chi-square value that its inspection produces is as shown in the left side; All cases are divided into blue and grey two groups (Y-axis represents case load for middle square chart, the expression mark of X-axis expressive notation thing) according to the expression of gene label by X-tile software; The right display Kaplan-Meier survivorship curve figure.
Form a team to arrange in training, the correlativity of scoring and small liver cancer patient RFS is expressed by X-tlie software analysis 30 gene labels, thus determine the cut off value of each gene label, be introduced into again internal verification group and separate outer checking group is verified according to the cut off value that training group obtains, the cut off value of each label and associated verification group result are as shown in table 2.3.
The cut off value that each gene label of table 2.3 is expressed and the result thereof
3.3 build the small liver cancer risk of recurrence model based on 6 gene labels
Form a team to arrange in training, inventor applies the prognostic model that LASSOCox regression model sets up small liver cancer recurrence.It is zero that LASSO uses L1 to reduce some regression coefficients to punishment.λ punishes parameter, and also claim tuning parameter, control the amount of contraction: λ value is larger, the predictor quantity of selection is fewer.LASSO recurrence has been expanded and has been widely used in the Cox proportional hazards regression models of survival analysis and high dimensional data.LASSO can be used for the optimal selection of higher-dimension gene microarray data, and these microarray datas are owing to having powerful prognosis meaning and the low correlation each other between variable can prevent overfitting.Inventor adopts punishment Cox regression model and LASSO to punish and reaches simultaneously and shrinks and the object of choice variable.Ten-time cross validation is used to determine the optimum value of λ.Inventor selects λ value by 1-standard error principle: best λ value is for the maximal value of those partial likelihood deviations in a standard error of partial likelihood deviation minimum value.In this research, inventor is drawn by partial likelihood deviation and log (λ), and λ is tuning parameter (Fig. 3).LASSOCox regression model establishes 6 and recurs relevant gene label to small liver cancer: comprise CD147, IL-7, Ki67, MIF, P97 and PD-L1 (Fig. 3).In Fig. 3, the biomarker LASSO coefficient distribution overview that (A) 30 liver cancer are relevant: what perpendicular line showed is the threshold value determined by 10 times of cross validations; (B) the Ten-time cross validation of the tuning Selecting parameter of LASSO model: solid vertical line represents partial likelihood deviation ± SE (standard error), the optimum value that vertical dotted line representative is obtained by minimum sandards and 1-SE standard.Inventor obtains the palindromia risk score that a formula based on 6 gene label expressions calculates each patient subsequently.According to patient's 6 gene label expressions separately, inventor applies the risk score that LASSOCox regression model calculates every patient: risk score=(0.03773100 × CD147 expression status)+(0.05809367 × IL-7 expression status)+(0.24175387 × Ki67 expression status)+(0.23151294 × MIF expression status)+(0.03817919 × P97 expression status)+(0.25708962 × PD-L1 expression status).At this formula, the height of each gene label is expressed cut off value and is utilized X-tile software to obtain, and the low expression status of each gene label equals 0, and high expressed state then equals 1.
3.4 detect and verify the risk of recurrence model based on 6 gene label sorters
The result of the best cut off value that comprehensive X-tile software produces and LASSOCox regretional analysis, relation between the distribution of the further evaluation and grading mark of inventor and survival of patients state, the result of inventor shows the lower patient of the risk score patient prognosis good (table 2.4 and Fig. 4 A) higher than risk score.Inventor with time dependent ROC curve assessment 6 gene label sorters in the accuracy (Fig. 4 A) of different follow up time prediction small liver cancer Patients on Recurrence.Clinicopathologic features be distributed in excessive risk and low-risk group not obviously different (table 2.2), but the high risk of recurrence crowd of 6 gene label sorters definition is 2.721 (95%CI:1.633-4.532 with the danger ratio of low risk of recurrence crowd; P<0.0001; Fig. 4 A).
Subsequently, inventor carries out same analysis also obtain similar result (Fig. 4 B) to internal verification row (167 example) of forming a team.In order to verify whether 6 gene label sorters have same prediction efficiency at different crowd, this sorter of inventor is applied to forms a team in row from the not concentric separate outer checking including 347 routine small liver cancers.Inventor finds: the high risk of recurrence crowd of sorter definition is 1.868 (95%CI:1.334-2.617 with the danger ratio of low risk of recurrence crowd; P<0.0001; Fig. 4 C).
By getting rid of the synergy of clinicopathologic features, it is the powerful independent prognostic factor affecting small liver cancer recurrence that multivariate survival analysis result shows 6 gene label sorters, no matter be in training group and internal verification group, or all obtain similar result (table 2.4) in external certificate group.Layering survival analysis result shows the prognostic model (Fig. 5 and Fig. 6) that 66 gene label sorters remain small liver cancer clinically and statistically significance.ROC tracing analysis shows 6 gene label sorters and have better accuracy (Fig. 7) than clinical risk factors in the recurrence of prediction small liver cancer.Therefore, the result of inventor shows that 6 gene label sorters can improve the prognostic value of clinicopathologic features.
As inventor result of study shown in, serum afp well can not predict the prognosis of small liver cancer patient.In order to clinician provides a quantivative approach to predict the possibility of AFP negative and positive group small liver cancer tumor recurrence, inventor constructs the alignment diagram that two integrate 6 gene label sorters and clinical high risk factor: AFP negative group and the positive group (Fig. 8) of AFP.Rule figure display column line chart is effect fine (Fig. 8) compared with ideal model.Time dependent ROC tracing analysis display column line chart has higher prediction accuracy (Fig. 8) the patient of AFP negative.
Table 2.4 is based on the single argument survival analysis of 6 gene label sorters, clinicopathologic features and individual gene label
* the median age.
Table 2.5 based on the sorter of 6 gene labels and small liver cancer patient without recurring the multivariable analysis of surviving
* the median age.
Conclusion
In this research, inventor uses and occurs to develop relevant genetic immunization group expression to hepatocellular carcinoma in the routine little HCC tumor tissues of LASSOCox analysis of regression model 335, identifies a sorter (comprising CD147, IL-7, Ki67, MIF, P97 and PD-L1) based on 6 genetic immunization group features.335 routine little HCC cases are randomized into a training group (n=168) and an internal verification group (n=167).In training group, inventor 6 of using LASSOCox regression model to filter out recur relevant gene label to small liver cancer, determine the prediction efficiency of 6 gene label sorters subsequently in internal verification group.In addition, inventor also have collected 347 patients's (individual authentication group) be treated surgically in other hospitals and does checking further to the result of inventor.Increasing sample size is also improve statistics usefulness in order to the false positive results reduced in screening process.The result of study of inventor is presented in little HCC tissue, the high risk of recurrence meaning patient Geng Gao of the dangerous integration of this 6 gene label sorter.The data set of this strengthening allows inventor can identify one to comprise 3 clinicopathologic features and 6 gene expression profiles, and for Accurate Prediction, which little HCC patient easily tumor recurrence occurs after excision, especially the patient that do not raise of those AFP levels.In this research, inventor uses the prognosis classification device based on biomarker-LASSO to verify this only containing the usefulness of the characteristic spectrum of 6 gene expressions.The use of LASSOCox regression model makes inventor can integrate multiple gene in an instrument, and this makes the accuracy of prognosis prediction be greatly enhanced than single-gene mark.Therefore, inventor thinks, it is reliable that this 6 allelic expression composes the data obtained, and is evaluate the strong prognostic model of small liver cancer Patients on Recurrence risk.
This 6 gene label sorter that inventor identifies has the potential possibility from scientific research to clinical practice, is particularly useful for little patient HCC identifying that those have high risk of recurrence.Although early stage HCC can by excision, transplant or local ablation therapy, the early stage HCC more than 60% is expert at after operative treatment can there is tumor recurrence.But, because the pathogenetic complicacy of HCC, be still difficult to the risk of predicting tumors recurrence after a resection at present.Inventor believes, is necessary very much clinically, when diagnosing, this kind of little patient HCC is divided into high risk of recurrence colony and low risk of recurrence colony.The object using this system to do to classify may be: do chemoprophylaxis targetedly or follow the tracks of at least more fully to follow up a case by regular visits to the little HCC of excessive risk.Inventor believes, this technology can be followed up a case by regular visits to by more positive tracking or early intervention improves the overall survival of little patient HCC; So this sorting technique may change the processing mode of little HCC medical treatment and nursing.According to the inventors knowledge, the recurrence of little HCC patient is only had a few biomarker to be in the news to can be used for predicting at present.Although these researchs are all very valuable, up to the present, also there is no a kind of multiple labeling forecast model based on large-scale multicenter study, for predicting the recurrence of the little HCC (HCC that especially that serum afp is not high) that HBV is relevant.In the research of inventor, this 6 gene label sorter assessment to little HCC risk of recurrence has independently clinical meaning, especially can not the small-sized early hepatocyte cancer of accurate evaluation to the grouping of those Present clinical.
In a word, small liver cancer patient effectively can be divided into height risk of recurrence group by the result of study prognostic evaluation system shown based on 6 allelic expression spectrums of inventor, integrates the predictive value that greatly can improve small liver cancer Index for diagnosis inside this categorizing system to traditional clinical pathology risk factors.According to the inventors knowledge, inventor studies first and demonstrates the gene expression characteristics spectrum that a group accurately can judge the little HCC prognosis of entity.Inventor find this organize 6 allelic expressions spectrums and be closely related with the postoperative risk of recurrence of little patient HCC.This mark can well for making tumor recurrence risk stratification to patient.
Claims (5)
1. combine mark for diagnosing small hepatic carcinoma, by CD147, IL-7, Ki67, MIF, P97 and PD-L1 totally 6 gene labels form.
2. gene label group is as the application of diagnosing small hepatic carcinoma mark, it is characterized in that: gene label group by CD147, IL-7, Ki67, MIF, P97 and PD-L1 totally 6 gene labels form.
3. detect the application of gene label group expression reagent as diagnosing small hepatic carcinoma recurrence reagent, it is characterized in that: gene label group by CD147, IL-7, Ki67, MIF, P97 and PD-L1 totally 6 gene labels form.
4. the application according to Claims 2 or 3, it is characterized in that: risk score=(0.03773100 × CD147 expression status)+(0.05809367 × IL-7 expression status)+(0.24175387 × Ki67 expression status)+(0.23151294 × MIF expression status)+(0.03817919 × P97 expression status)+(0.25708962 × PD-L1 expression status), at this formula, the height of each gene label is expressed cut off value and is utilized X-tile software to obtain, be equal to or less than cut off value and be decided to be low expression, high expressed is called higher than cut off value, low expression status equals 0, high expressed state then equals 1.
5. application according to claim 3, is characterized in that: detecting gene label group expression reagent is six antibody such as anti-human CD147, IL-7, Ki67, MIF, P97 and PD-L1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510744435.1A CN105319364B (en) | 2015-10-28 | 2015-11-03 | For predicting that the Combining diagnosis of microhepatia cancer recurrence is marked |
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510718239 | 2015-10-28 | ||
| CN2015107182397 | 2015-10-28 | ||
| CN201510744435.1A CN105319364B (en) | 2015-10-28 | 2015-11-03 | For predicting that the Combining diagnosis of microhepatia cancer recurrence is marked |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105319364A true CN105319364A (en) | 2016-02-10 |
| CN105319364B CN105319364B (en) | 2017-09-05 |
Family
ID=55247210
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510744435.1A Active CN105319364B (en) | 2015-10-28 | 2015-11-03 | For predicting that the Combining diagnosis of microhepatia cancer recurrence is marked |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105319364B (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107463798A (en) * | 2017-08-02 | 2017-12-12 | 南京高新生物医药公共服务平台有限公司 | Predict the 12 gene expressions classification device and its construction method of adenocarcinoma of colon prognosis |
| CN107561280A (en) * | 2017-09-30 | 2018-01-09 | 四川大学华西医院 | A kind of kit for predicting breast cancer relapse |
| CN107653318A (en) * | 2017-10-19 | 2018-02-02 | 中山大学肿瘤防治中心 | One group is used for the mark and its application for predicting that nasopharyngeal carcinoma shifts risk |
| CN110634571A (en) * | 2019-09-20 | 2019-12-31 | 四川省人民医院 | Prognosis prediction system after liver transplantation |
| CN111088352A (en) * | 2019-11-28 | 2020-05-01 | 浙江大学 | Establishment method and application of polygenic liver cancer prognosis grading system |
| CN112329876A (en) * | 2020-11-16 | 2021-02-05 | 中山大学附属第六医院 | Colorectal cancer prognosis prediction method and device based on image omics |
| CN112831562A (en) * | 2021-01-25 | 2021-05-25 | 浙江科技学院 | Biomarker combination and kit for predicting recurrence risk of liver cancer patient after resection |
| CN113436741A (en) * | 2021-07-16 | 2021-09-24 | 四川大学华西医院 | Lung cancer recurrence prediction method based on tissue specific enhancer region DNA methylation |
| CN113488108A (en) * | 2021-07-02 | 2021-10-08 | 温州医科大学 | Novel model for predicting individual recurrence risk of acute anterior uveitis and application thereof |
| CN115862875A (en) * | 2023-02-27 | 2023-03-28 | 四川大学华西医院 | Postoperative pulmonary complication prediction method and system based on multi-type feature fusion |
-
2015
- 2015-11-03 CN CN201510744435.1A patent/CN105319364B/en active Active
Non-Patent Citations (5)
| Title |
|---|
| 刘倩等: "MIF和CD147在非小细胞肺癌中的表达及意义", 《四川大学学报》 * |
| 明健等: "IL-7/IL-7R在非小细胞肺癌中的表达及与淋巴转移和预后的关系", 《中国肺癌杂志》 * |
| 杨兆红等: "非小细胞肺癌中p97蛋白的表达及其临床意义", 《医学创新研究》 * |
| 王旭晖等: "Ki67、bcl-2、bax蛋白在非小细胞肺癌中的表达及预后价值", 《首都医科大学学报》 * |
| 马薇等: "PD-L1和PD-1在非小细胞肺癌中的表达及其临床意义", 《实用医学杂志》 * |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107463798A (en) * | 2017-08-02 | 2017-12-12 | 南京高新生物医药公共服务平台有限公司 | Predict the 12 gene expressions classification device and its construction method of adenocarcinoma of colon prognosis |
| CN107561280A (en) * | 2017-09-30 | 2018-01-09 | 四川大学华西医院 | A kind of kit for predicting breast cancer relapse |
| CN107561280B (en) * | 2017-09-30 | 2019-07-26 | 四川大学华西医院 | A kind of kit for predicting breast cancer relapse |
| CN107653318A (en) * | 2017-10-19 | 2018-02-02 | 中山大学肿瘤防治中心 | One group is used for the mark and its application for predicting that nasopharyngeal carcinoma shifts risk |
| CN107653318B (en) * | 2017-10-19 | 2020-02-21 | 中山大学肿瘤防治中心 | Marker for predicting nasopharyngeal carcinoma metastasis risk and application thereof |
| CN110634571A (en) * | 2019-09-20 | 2019-12-31 | 四川省人民医院 | Prognosis prediction system after liver transplantation |
| CN111088352A (en) * | 2019-11-28 | 2020-05-01 | 浙江大学 | Establishment method and application of polygenic liver cancer prognosis grading system |
| CN111088352B (en) * | 2019-11-28 | 2022-02-08 | 浙江大学 | Establishment method and application of polygenic liver cancer prognosis grading system |
| CN112329876A (en) * | 2020-11-16 | 2021-02-05 | 中山大学附属第六医院 | Colorectal cancer prognosis prediction method and device based on image omics |
| CN112831562A (en) * | 2021-01-25 | 2021-05-25 | 浙江科技学院 | Biomarker combination and kit for predicting recurrence risk of liver cancer patient after resection |
| CN113488108A (en) * | 2021-07-02 | 2021-10-08 | 温州医科大学 | Novel model for predicting individual recurrence risk of acute anterior uveitis and application thereof |
| CN113436741A (en) * | 2021-07-16 | 2021-09-24 | 四川大学华西医院 | Lung cancer recurrence prediction method based on tissue specific enhancer region DNA methylation |
| CN113436741B (en) * | 2021-07-16 | 2023-02-28 | 四川大学华西医院 | Lung cancer recurrence prediction method based on tissue specific enhancer region DNA methylation |
| CN115862875A (en) * | 2023-02-27 | 2023-03-28 | 四川大学华西医院 | Postoperative pulmonary complication prediction method and system based on multi-type feature fusion |
| CN115862875B (en) * | 2023-02-27 | 2024-02-09 | 四川大学华西医院 | Postoperative pulmonary complications prediction method and system based on multi-type feature fusion |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105319364B (en) | 2017-09-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105319364A (en) | Combined diagnosis marker for predicting small hepatocellular carcinoma relapse | |
| Attard et al. | Duplication of the fusion of TMPRSS2 to ERG sequences identifies fatal human prostate cancer | |
| Boers et al. | HER2 status in gastro‐oesophageal adenocarcinomas assessed by two rabbit monoclonal antibodies (SP3 and 4B5) and two in situ hybridization methods (FISH and SISH) | |
| US20030190602A1 (en) | Cell-based detection and differentiation of disease states | |
| Geradts et al. | The oncotype DX recurrence score is correlated with a composite index including routinely reported pathobiologic features | |
| JP2004526154A (en) | Cell-based detection and disease state discrimination | |
| Lin et al. | High-plex immunofluorescence imaging and traditional histology of the same tissue section for discovering image-based biomarkers | |
| Sebo | What are the keys to successful thyroid FNA interpretation? | |
| Ye et al. | Circulating tumor cells as a potential biomarker for postoperative clinical outcome in HBV-related hepatocellular carcinoma | |
| CN109897899A (en) | A kind of marker and its application for Locally Advanced esophageal squamous cell carcinoma Index for diagnosis | |
| Sigel et al. | Predicting pulmonary adenocarcinoma outcome based on a cytology grading system | |
| Khoddami et al. | Correlation between Gleason scores in needle biopsy and corresponding radical prostatectomy specimens: a twelve-year review | |
| CN106248945A (en) | Patients with hepatocellular carcinoma is carried out method, system and the test kit of the packet of hepatocarcinoma radical excision prognosis situation | |
| Chung et al. | Clinical and technical guideline for endoscopic ultrasound (EUS)-guided tissue acquisition of pancreatic solid tumor: Korean Society of Gastrointestinal Endoscopy (KSGE) | |
| Soave et al. | Detection and oncological effect of circulating tumour cells in patients with variant urothelial carcinoma histology treated with radical cystectomy | |
| Bose et al. | Fractal analysis of nuclear histology integrates tumor and stromal features into a single prognostic factor of the oral cancer microenvironment | |
| Nieva et al. | Fluid biopsy for solid tumors: a patient’s companion for lifelong characterization of their disease | |
| Barr Fritcher et al. | FISH ing for pancreatobiliary tract malignancy in endoscopic brushings enhances the sensitivity of routine cytology | |
| Sanguedolce et al. | Evolving concepts and use of immunohistochemical biomarkers in flat non-neoplastic urothelial lesions: WHO 2016 classification update with diagnostic algorithm | |
| Shim et al. | Diagnostic algorithm for papillary urothelial tumors in the urinary bladder | |
| Connolly et al. | Role of the Surgical Pathologist in the Diagnosis and Management of the Cancer Patient | |
| CN106290874B (en) | Method, system and the kit of the grouping of transplantation of liver prognosis situation are carried out to patients with hepatocellular carcinoma | |
| CN105986024A (en) | Genes for prognosis of triple negative breast cancer and application thereof | |
| Ibrahim et al. | Improving mitotic cell counting accuracy and efficiency using phosphohistone‐H3 (PHH3) antibody counterstained with haematoxylin and eosin as part of breast cancer grading | |
| Zhu et al. | Identification and characterization of effusion tumor cells (ETCs) from remnant pleural effusion specimens |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |