CN111965239A

CN111965239A - Products, uses and methods for gastric cancer-related screening and assessment

Info

Publication number: CN111965239A
Application number: CN201910420649.1A
Authority: CN
Inventors: 钟晟; 郑杰; 张华�
Original assignee: Shanghai Changtai Biotechnology Co ltd
Current assignee: Shanghai Changtai Biotechnology Co ltd
Priority date: 2019-05-20
Filing date: 2019-05-20
Publication date: 2020-11-20

Abstract

The invention provides a product, application and a method for Gastric Cancer (GC) related screening and evaluation, which are characterized in that body fluid samples of Gastric cancer patients and healthy people are mixed with nanosphere materials, metabolic molecular information in related body fluid is collected through matrix assisted laser desorption ionization time of flight mass spectrometry (MALDI-TOF-MS), the metabolic molecular spectrum difference of the healthy people and Gastric cancer patients is distinguished by an artificial intelligence classification algorithm, and the difference is classified, so that the distinguishing effect of high accuracy, high specificity and high sensitivity is achieved, or the product becomes a tool for Gastric cancer related screening and evaluation.

Description

Products, uses and methods for gastric cancer-related screening and assessment

Technical Field

The invention relates to the fields of biology and doctors, in particular to a product, application and a method for screening and evaluating related gastric cancer.

Background

Gastric Cancer (GC) is a malignant tumor originated from Gastric mucosal epithelium, has obvious regional characteristics, and is mainly developed in northwest and southeast coastal areas of China. The precancerous lesion of the stomach mainly comprises abnormal hyperplasia of gastric mucosa and intestinal metaplasia, and is mainly accompanied with chronic atrophic gastritis, namely the occurrence of gastric cancer is a multi-step process, namely chronic superficial gastritis-atrophic gastritis-intestinal metaplasia-dysplasia-gastric cancer. In developed countries such as japan, europe, and the united states, the prevalence of gastric cancer has been greatly reduced, but in china, gastric cancer is still the first five high-incidence tumors. Gastroscope is the most common means of clinical stomach cancer screening, but because the gastroscope is invasive, the examiner needs local anesthesia even general anesthesia, leads to gastroscope compliance greatly reduced, and the chinese crowd lacks the consciousness of regularly examining the gastroscope. For early screening of gastric cancer, a noninvasive, accurate, simple and high-cost-performance product is in urgent need at present, and the mixed liquid prepared from the nano material and the human body liquid is used for MALDI-TOF-MS detection, so that efficient gastric cancer screening in large-scale people can be realized, or the product becomes a powerful tool for early screening of gastric cancer.

Disclosure of Invention

The invention aims to solve the technical problem of the prior art and provides a product, application and a method for screening and evaluating related gastric cancer.

In order to achieve the purpose, the technical scheme adopted by the invention is as follows:

a product for gastric cancer screening, early diagnosis, risk assessment, the product being an operable system, the product comprising:

(S1) the nanometer material used for selectively enhancing the resolution of the metabolic molecule of the measured body fluid, and the matrix-assisted laser desorption ionization time-of-flight mass spectrometer used for measuring the metabolic molecule signal of the body fluid are used for measuring different metabolic signals of the body fluids of patients with different sources and healthy people through a core artificial intelligence classification algorithm;

(S2), a module and a processor for calculating the ratio of metabolic molecular composition and content of healthy persons to gastric cancer;

(S3), optionally, a module and a processor for determining whether gastric cancer is present, and for assessing the risk of developing a tumor in a subject based on the composition and content of the metabolic molecules in the body fluid.

In the product for screening, early diagnosis and risk assessment of gastric cancer, the body fluid is selected from serum, plasma, saliva and urine.

In the product for screening, early diagnosis and risk assessment of gastric cancer provided by the invention, the reagent for determining body fluid metabolism in the step (S1) is a nano gold ball material, a matrix-assisted laser desorption ionization time-of-flight mass spectrum and a corresponding matched target plate.

The product for screening, early diagnosis and risk assessment of gastric cancer provided by the invention further comprises one or more selected from the following group:

a. reagents and instruments for collecting and processing a sample of a subject's bodily fluid;

b. reagents and instruments for separating and purifying serum/or plasma/or saliva/or urine;

c. reagents and instruments for separating and purifying serum/plasma/saliva/urine;

d. reagents and instruments for separating, purifying, enriching serum/plasma/saliva/urine;

e. the system comprises a database, a module and a processor for storing and processing metabolic molecular indexes of serum/plasma/saliva/urine of a subject;

f. a module and a processor for providing a decision threshold;

g. a module and a processor for comparing the metabolic molecular profile of the subject to a control to determine whether the subject suffers from gastric cancer, to make a prognostic assessment or risk assessment of suffering from gastric cancer, to monitor a condition or to make a medical assessment;

h. a module and processor for providing a diagnosis, test result report;

j. instructions or instructions for use in which one or more of the following applications and decisions are described:

(i) for gastric cancer screening test: when all the detected molecular weights are 100-3000 daltons and the ratio between the number of metabolic molecules with a signal-to-noise ratio greater than 3.05 reaches or is higher than a preset threshold value, determining that the subject has gastric cancer or is at higher risk of having gastric cancer;

(ii) for judgment of health and malignancy: when all detected molecular weights are 100-3000 daltons and the ratio of the number of the metabolic molecules with the signal-to-noise ratio larger than 2.9 reaches or is higher than a preset threshold value, judging that the subject has gastric cancer or is at risk of having gastric cancer;

(iii) for prognostic monitoring: when the detected molecular weights are 100-3000 daltons and the ratio of the number of the metabolic molecules with the signal-to-noise ratio larger than 2.65 reaches or is higher than a preset threshold value, the subject is prompted to relapse gastric cancer or is at risk of suffering from gastric cancer;

(iv) prediction of gastric cancer occurrence in a subject population: when the detected molecular weights are 100-;

(v) for monitoring the condition of gastric cancer patients: when the detected molecular weight is 100-;

(vi) for efficacy assessment: the ratio of the number of all detected metabolic molecules with the molecular weight of 100-3000 daltons and the signal-to-noise ratio of more than 2.55 is measured and calculated at different time points before and after diagnosis and treatment or in the treatment process, and when the difference of the result higher than the result obtained by the previous measurement reaches or is higher than a preset range, the treatment effect of the therapy or the drug is poor.

In the products for screening, early diagnosis and risk assessment of gastric cancer, the preset threshold values in (i) - (iv) are determined by the optimal threshold values of 2.35-3.05, wherein the optimal threshold values are determined based on an artificial intelligence classification algorithm.

In the product for screening, early diagnosis and risk assessment of gastric cancer, the preset threshold value in (i) - (iv) is 2.35-3.05.

In the product for screening, early diagnosis and risk assessment of gastric cancer provided by the invention, in the step (vi), when the number of the metabolic molecules with the signal-to-noise ratio of the detected molecular signal of more than 2.55 is higher than that obtained in the previous test, the condition of the patient is shown to be developed or further worsened.

In the product for screening, early diagnosis and risk assessment of gastric cancer provided by the invention, the ratio calculation of the number of molecules with signal-to-noise ratio higher than the optimal threshold value and the number of the selected detected metabolic molecules is carried out by adopting the following formula:

score ratio-number of molecules with signal to noise ratio above the optimal threshold/number of detected metabolic molecules.

The product for screening, early diagnosis and risk assessment of gastric cancer is used in a detection method comprising the following steps:

1. determining the level of a metabolic molecule in a plasma/serum/saliva/urine sample of a subject, said level of a metabolic molecule comprising: a metabolic molecule having a molecular weight of 100-;

2. calculating the ratio of the number of molecules with signal-to-noise ratio higher than the optimal threshold value to the number of detected metabolic molecules;

3. determining whether the subject has gastric cancer based on the level of the metabolic molecule, performing a prognostic assessment or risk assessment of the subject having gastric cancer, monitoring the condition of the subject, or assessing the efficacy of the treatment.

In the product for screening, early diagnosis and risk assessment of gastric cancer provided by the invention, the method further comprises one or more of the following steps:

1) collecting and processing a subject plasma/serum/saliva/urine sample;

2) separating and purifying serum/plasma/saliva/urine;

3) separating and purifying serum/plasma/saliva/urine;

4) separating, purifying, enriching serum/plasma/saliva/urine;

5) the metabolic molecule content ratio of the serum of the storage and processing object is used as an index of the gastric cancer related screening;

6) providing a judgment threshold value;

7) comparing the metabolic molecular composition content of the subject with a control to determine whether the subject suffers from gastric cancer, performing prognosis evaluation or suffers from gastric cancer, monitoring the condition of the subject or evaluating the curative effect of the subject;

8) providing diagnosis and detection result report;

9) monitoring other gastric cancer indexes, and combining the detection results of the other gastric cancer indexes with the disease condition monitoring result.

The product for screening, early diagnosis and risk assessment of gastric cancer also comprises a kit, a device, a system and a combination thereof for detecting other gastric cancer indexes.

In the product for screening, early diagnosis and risk assessment of gastric cancer provided by the invention, the index is selected from the following indexes: metabolic molecules of serum/plasma/saliva/urine of a patient.

In the product for screening, early diagnosis and risk assessment of gastric cancer provided by the invention, the basic condition of the patient is selected from the following: the number of molecules whose metabolic molecular signal-to-noise ratio is above an optimal threshold.

In the product for screening, early diagnosis and risk assessment of gastric cancer, the clinical manifestations of the tumor are selected from one or more of the following: complete health/abnormal intestinal metaplasia/definitive diagnosis of gastric cancer/gastric cancer cure/gastric cancer recurrence.

In the product for screening, early diagnosis and risk assessment of gastric cancer provided by the invention, the specific physicochemical examination index comprises a tumor biochemical index and medical image data, wherein the tumor biochemical index is selected from one or more of the following: CEA, CA19-9, CA242, CA50, PGI, PGII in plasma.

Use of an apparatus, module and processor for the manufacture of a product for gastric cancer screening, early diagnosis, prognosis evaluation, risk assessment, condition monitoring and efficacy assessment of a subject, the product being a device, system and combination thereof, the apparatus, module and processor comprising:

A. reagents, instruments, modules and processors for determining the levels of serum/plasma/saliva/urine metabolic molecules for determining the composition and content of metabolic molecules;

B. optionally, a module or processor for calculating that the signal-to-noise ratio of the metabolic molecule is above a certain threshold;

C. optionally, a module or processor for determining whether the subject suffers from gastric cancer, performing a prognostic assessment or risk assessment of suffering from gastric cancer, monitoring the condition of the subject, or assessing the efficacy of the treatment based on the ratio of the signal to noise ratios.

The instrument, the module and the processor are applied to preparing products for gastric cancer screening, early diagnosis, prognosis evaluation, risk evaluation, disease monitoring and curative effect evaluation of a subject, wherein the instrument, the module and the processor are used for determining blood metabolites, and after the blood metabolites are combined with a nanogold ball material, matrix-assisted laser desorption ionization time-of-flight mass spectrometry is used for measuring signal-to-noise ratio data of the sample metabolites, and then a disease performance judgment and evaluation result is given based on an artificial intelligence algorithm.

Compared with the prior art, the invention has the advantages that: the invention develops a new method for diagnosing and typing gastric cancer based on a small amount or trace (less than 5 mL) of body fluid samples (including plasma, serum, urine, saliva and the like). The accuracy (sensitivity and specificity) of the method for diagnosing the gastric cancer is higher than that of the current clinical gold standard, the early gastric cancer with intestinal metaplasia and genetic hyperplasia difficult to judge in the aspect of imaging can be identified, and the method has prompting significance for the benign and malignant tumors and the pathogenic gene mutation carried by the tumors. In addition, compared with the conventional gastroscope index detection, the mass spectrum technology adopted in the invention has the advantages of shorter time consumption, no harm to human bodies, more convenient detection and greatly reduced cost.

Drawings

FIG. 1 is a schematic view of a sample collection device.

FIG. 2 is a schematic view of a sample storage device.

FIG. 3 is a schematic view of a sample preparation device.

FIG. 4 is a schematic diagram of a mass spectrometry data acquisition device.

FIG. 5 is a schematic diagram of the data generated by the mass spectrometric data acquisition apparatus.

FIG. 6 is a schematic diagram of a classification analysis of mass spectral data using an artificial intelligence algorithm.

FIG. 7 is a schematic diagram of the biological signal pathway for identifying relevant metabolite molecules that can be used for basic classification and studying the association of the metabolite molecules.

Fig. 8 is a flowchart of an algorithm for diagnosing and typing gastric cancer and predicting which disease-causing gene mutation is carried by a tumor.

Detailed Description

The technical solution adopted by the present invention will be further explained with reference to the schematic drawings.

The invention needs to obtain a high-accuracy and high-sensitivity prediction classification model which is based on a large number of real biological samples and can be used for repeated detection application, and the specific method comprises the following steps:

1. mixing a biological sample (blood plasma, blood serum, urine, saliva or the like) with a nano material for selectively enhancing the resolution of measuring the metabolism molecules of the body fluid, wherein the nano material comprises but is not limited to silicon and silicon compounds, gold, silver, platinum and other noble metals;

2. preparing the mixed sample on a target plate of a mass spectrum;

3. obtaining data (with accuracy of 0.001-0.1 Dalton) by using a mass spectrometer, wherein the obtained data is relative number value of the metabolic molecules with the length of more than 100 Dalton, and the data of all the metabolic molecules with the length of 100 Dalton and the signal-to-noise ratio of more than 2.4 are selected;

4. grouping and classifying the obtained data according to requirements, wherein the classification mode is as follows: diseased Vs healthy, metastatic Vs non-metastatic, types of gene mutations, etc.;

5. the data is classified using a machine learning algorithm, typically trained using 70% of the data taken randomly, tested using 30% of the data, and the model accuracy is cross-validated, where the purpose of data training and validation is: a high accuracy, high sensitivity predictive classification model based on a large number of real biological samples is obtained that can be used for repeated detection applications.

FIG. 1 represents the collection of serum or plasma or saliva or urine from a test person using a sample collection device; FIG. 2 is a schematic representation of the storage of collected samples using a sample storage device, the storage of collected samples at-80 ℃ and the transport to the corresponding detection laboratory; FIG. 3 is a schematic diagram of a sample preparation apparatus for mixing a sample with prepared nanomaterial to prepare the sample before processing; FIG. 4 is a schematic view of a mass spectrometry data acquisition device to obtain mass spectrometry data of a sample; FIG. 5 is a schematic diagram of the data generated by the mass spectrometric data acquisition device, which is a graph of the data generated by the mass spectrometric data acquisition device; FIG. 6 is a schematic diagram of a classification analysis of mass spectrometry data using an artificial intelligence algorithm; FIG. 7 is a schematic diagram of the biological signal pathway for identifying relevant metabolite molecules that can be used for basic classification and studying the association of the metabolite molecules. Fig. 1-7 are experimental and analytical routes for screening and assessment related to gastric cancer. Fig. 8 is a flowchart of an algorithm for diagnosing and typing gastric cancer and predicting which pathogenic mutations are carried by tumors, which includes a data preprocessing module and a hyper-parameter selection module, and the details of which will be described later.

The invention utilizes MALDI-TOF-MS (Matrix Assisted Laser Desorption/Ionization Time of Flight Mass Spectrometry) based on a new material to detect a large number of clinically confirmed gastric cancer patients and plasma of non-gastric cancer control population at high flux, and uses the obtained Mass data for machine learning.

The technology of the invention adopts laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS) based on nanometer results (including but not limited to quantum dots, nanowires, nanospheres, nano hollow spheres, nano triangular shapes and the like) of nanometer materials (including but not limited to silicon and silicon compounds, gold, silver, platinum and other precious metals) to detect metabolites (substances with molecular weight below 3000 daltons) of plasma samples of a large number of gastric cancer patients who have been diagnosed, and a set of brand-new algorithms for diagnosing and typing gastric cancer and predicting which pathogenic gene mutation is carried by tumor are developed by learning big data by using a machine, wherein the overall flow chart of the algorithm is shown in FIG. 8, and the specific contents of the algorithm are as follows:

1) converting the raw data derived from the mass spectrometer (i.e. the aforementioned "relative number value of metabolic molecules above 100 daltons, we select 100-;

2) repeating the grouping of the data set processed in the step 1) for 10 times, wherein 30% of the groups are used as a test data set and 70% of the groups are used as a model training data set. A layered sampling method is adopted during grouping, so that the distribution of positive and negative samples in each group of data is consistent with the distribution of positive and negative samples in the total samples;

3) denoising and smoothing the data set processed in the step 2);

4) carrying out baseline calibration operation on the data set processed in the step 3);

5) carrying out standardization operation on the data set processed in the step 4) to obtain an integral mean value and variance parameters of the training set data;

6) aligning the data set processed in the step 5);

7) and repeating five-fold cross validation for the model training set data in the step 6) for 5 times, wherein the data are divided into a validation data set of 20% and a training data set of 80%. A layered sampling method is adopted during grouping, so that the distribution of positive and negative samples in each group of data is consistent with the distribution of positive and negative samples in the total samples;

8) training the data in the step 7) by using a classifier, outputting the average verification precision of the classifier model, and selecting the hyper-parameter of the classifier model when the verification precision is highest. The classifier model specifically adopts a Support Vector Machine (SVM), the SVM is a machine learning model utilizing hyperplane segmentation data, the SVM is generally applied to a two-classification problem under a supervised learning situation, a learning strategy is interval maximization, and SVM solution can be converted into a convex quadratic programming problem. The linear SVM is proposed by cortex and Vapnik in 1995, has great advantages in solving the problem of high-dimensional small sample classification, and can be popularized and applied to other machine learning problems such as function fitting and the like.

The idea of SVM classification is as follows: a hyperplane is found to segment the data (a straight line in the two-dimensional case) and then the segmentation of the data is maximized using the support vectors. The mathematical model is as follows:

simultaneously, the following requirements are met: yi (w)^T·x_i+b)-1≥0，i＝1，2，…，n

Where the sample points of the data set T are hyperplanes. If the dataset T is linearly separable, then a maximum spaced separation hyperplane exists and is unique that can completely separate the samples in the dataset.

9) Putting the whole model training data set into the classifier model with the highest precision and the hyperparameter in the step 8) for training to obtain a final classifier model;

10) denoising and smoothing the model test set data obtained in the step 2);

11) carrying out baseline removal work on the data set processed in the step 10);

12) carrying out standardization operation on the data set processed in the step 11) by using the standardization parameters of the model training set data obtained in the step 5);

13) aligning the data set processed in the step 12);

14) putting the data set processed in the step 13) into the final classifier model in the step 9) for testing;

15) obtaining a predicted value of each sample in the test data set through the processing of the classifier in the step 14);

16) comparing the predicted value obtained in the step 15) with the sample gold standard to obtain the average test error of the classifier model

Steps 1) to 16) are detailed for the "high accuracy, high sensitivity predictive classification models based on large numbers of real biological samples that can be used for repeated detection applications" obtaining method.

Example 1: screening/screening of gastric cancer by serum metabolic molecule time-of-flight mass spectrometry

The method comprises the following steps of determining the content of 100-3000 Da metabolic molecules in a control serum sample and a gastric cancer patient serum sample by adopting a nano gold ball material and a method, wherein the detection results are as follows:

1. by observing the metabolic molecular spectrogram of 100-3000 Da of serum samples of 500 patients in total, the number of the metabolic molecules with the signal-to-noise ratio of more than 3.05 can be found to be obviously different between gastric cancer patients and controls.

2. Comparing the metabolic molecular maps of 500 healthy human serum samples of 100-3000 Da, and simultaneously carrying out comparison analysis on the number of metabolic molecules with the signal-to-noise ratio of a test subject being more than 3.05. The result shows that the corresponding sensitivity is up to 91.6% under the condition that the specificity is controlled to be 98.5% by using the metabolic molecular number of 100-3000 Da, wherein the metabolic molecular number of which the signal-to-noise ratio is more than 3.05, and the total detected metabolic molecular number as the gastric cancer screening index. The index is highly accurate as a diagnosis index by referring to the judgment standard of the puncture tissue pathology.

3. Numerically, the difference between the metabolic molecules with the signal-to-noise ratio of more than 3.05 in the 100-3000 Da metabolic molecules of the gastric cancer patient and the control sample is very significant.

4. Meanwhile, the threshold of 3.05 is used as a criterion for judging that the subject has gastric cancer or is at risk of having gastric cancer, and the results are found in the following table 1:

TABLE 1

In the samples, the detection rate of the gastric cancer can reach about 91.6 percent;

the results show that the matrix-assisted laser desorption ionization time-of-flight mass spectrometry can be effectively used for screening gastric cancer with high sensitivity and high specificity, and compared with the existing standard, the method and the index have higher accuracy. Furthermore, the methods and indices of the present invention may be used in combination with existing criteria to further improve the detection rate of gastric cancer.

The above description is only a preferred embodiment of the present invention, and does not limit the present invention in any way. It will be understood by those skilled in the art that various changes, substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims.

Claims

1. A product for gastric cancer screening, early diagnosis, risk assessment, the product being an operable system, the product comprising:

(S2), a module and/or processor for calculating the ratio of metabolic molecular composition and content of healthy persons to gastric cancer patients;

(S3), optionally, a module and a processor for determining whether gastric cancer is present, and for assessing the risk of developing a tumor in a subject based on the composition and content of metabolic molecules in the body fluid;

(S4), optionally, the product further comprises:

(iv) prediction of gastric cancer onset for a subject population: when the detected molecular weights are 100-;

2. The product of claim 1, further comprising one or more selected from the group consisting of:

b. reagents and instruments for separating and purifying serum/plasma/saliva/urine;

f. a module and a processor for providing a decision threshold;

h. a module and a processor for providing diagnosis and detection result report.

3. The product of claim 1, wherein the predetermined threshold values in (i) - (iv) are determined by an optimal threshold value of 2.35-3.05, wherein the optimal threshold value is determined based on an artificial intelligence classification algorithm.

4. The product of claim 1, wherein (vi) a higher number of metabolic molecules than the previous test with a signal-to-noise ratio of 2.55 for the detected molecular signal indicates that the patient's condition has progressed or has further worsened.

5. The product of claim 1, wherein the calculation of the ratio of the number of molecules having a signal-to-noise ratio above the optimal threshold to the number of selected detected metabolic molecules is performed using the formula:

6. The product of claim 1, wherein the product is used in a test method comprising the steps of:

7. The product of claim 6, wherein the method further comprises one or more of the following:

1) collecting and processing a subject plasma/serum/saliva/urine sample;

2) separating and purifying serum/plasma/saliva/urine;

3) separating and purifying serum/plasma/saliva/urine;

4) separating, purifying, enriching serum/plasma/saliva/urine;

5) storing and processing the metabolic molecule content ratio of the serum/plasma/saliva/urine of the object as the index of the gastric cancer related screening;

6) providing a judgment threshold value;

8) providing diagnosis and detection result report;

8. The product of claim 1, further comprising kits, devices, systems, and combinations thereof for detecting other gastric cancer markers.

9. The product of claim 8, wherein the indicator is selected from the group consisting of: metabolic molecules of serum/plasma/saliva/urine of the patient;

the patient's basic condition is selected from the following: the number of molecules whose metabolic molecular signal-to-noise ratio is above an optimal threshold;

the clinical presentation of the tumor is selected from one or more of the following: complete health/abnormal intestinal metaplasia/definitive diagnosis of gastric cancer/gastric cancer cure/gastric cancer recurrence;

the specific physicochemical examination index comprises a tumor biochemical index and a medical image index, wherein the tumor biochemical index is selected from one or more of the following indexes: PGI, PGII, CEA in plasma.

10. Use of an apparatus, module and processor for the manufacture of a product for gastric cancer screening, early diagnosis, prognosis evaluation, risk assessment, condition monitoring and efficacy assessment of a subject, wherein the product is a device, system and combination thereof, the apparatus, module and processor comprising:

C. optionally, a module or processor for determining whether the subject suffers from gastric cancer, performing a prognostic assessment or risk assessment of suffering from gastric cancer, monitoring the condition of the subject, or assessing the efficacy of the treatment based on the ratio of the signal to noise ratios;

and (A) after the blood metabolite is combined with the nano gold ball material, matrix-assisted laser desorption ionization time-of-flight mass spectrometry is used for determining the blood metabolite, the signal-to-noise ratio data of the sample metabolite is measured, and then a disease performance judgment and evaluation result is given based on an artificial intelligence algorithm.