Prognostic risk assessment model of multiple myeloma as well as construction method and application of prognostic risk assessment model

Abstract

The invention relates to the technical field of clinical medicine, in particular to a prognosis risk assessment model of multiple myeloma as well as a construction method and application of the prognosis risk assessment model. Compared with the prior art, the prognosis risk assessment model for multiple myeloma provided by the invention has the advantages of wider coverage, higher prediction accuracy and higher layering ability. The sample data of the model is from a large queue of newly-diagnosed multiple myeloma patients of multiple medical centers in China, so that compared with other similar international models, the model is more in line with clinical, biological and prognosis characteristics of Chinese people, and has greater advantages in Chinese people.

Description

Technical Field

[0001] The present invention relates to the field of clinical medical technology, and in particular to a prognostic risk assessment model for multiple myeloma and a construction method and application thereof. Background Art

[0002] Multiple myeloma (MM) is a highly heterogeneous plasma cell malignancy and the second most common hematological malignancy, accounting for about 10% of all hematological malignancies. MM has highly unstable genetic characteristics, often manifested as obvious and variable chromosomal abnormalities, and is currently still an incurable disease. With the emergence of new therapeutic drugs and combined treatment regimens, the survival outcomes of MM patients have been significantly improved, but the survival outcomes of patients are still highly heterogeneous, ranging from a few months to more than ten years, even in the clinical trial population with unified treatment. The vast majority of MM patients have chromosomal abnormalities, including abnormalities in number and structure, abnormalities in gene expression, etc. These abnormalities will gradually accumulate during the course of the disease, thereby enabling MM cells to acquire malignant proliferation, invasiveness and drug resistance, which will inevitably lead to greater differences in disease progression and prognosis among different patients. Therefore, prognostic stratification is needed to formulate targeted treatment plans and judge prognosis.

[0003] With the continuous development of new detection technologies, new abnormalities are constantly being discovered, providing new insights into the biological and genetic characteristics of MM in a more comprehensive and systematic way, and providing a theoretical basis for more potential, more effective and safer treatment options for MM patients. One of the most important advances is the recognition of the genetic heterogeneity of MM and its important impact on patient prognosis. Combined with the prognostic factors discovered in previous studies, it provides the possibility for risk stratification and personalized treatment of MM patients. Understanding and controlling the accessibility of these drugs and their differences in different MM risk groups will further enhance people's ability to design treatment strategies based on risk stratification.

[0004] Looking at the development of the prognostic stratification system for MM, it can be seen that a variety of detection techniques, including fluorescence in situ hybridization, multi-parameter flow cytometry, gene expression profiling, SNP-array, etc., have played important roles in it. They can identify more molecular markers related to MM diagnosis and prognosis, making the MM prognostic stratification system more and more perfect. The staging system (DS) based on tumor burden proposed by Durie and Salmon in 1975 was the earliest prognostic stratification system for MM, mainly based on indicators such as hemoglobin, bone destruction, M protein amount, and serum calcium of patients, with great limitations. For example, in non-secretory and oligosecretory MM, the amount of M protein does not completely correlate with tumor burden. In addition, the application of new therapeutic drugs and combined regimens has greatly reduced the tumor burden of patients and also increased the inconsistency between DS staging and patient survival. Subsequently, in 2005, the International Myeloma Working Group (IMWG) analyzed the data of MM patients from 17 centers around the world and found that the levels of β2-MG and albumin were closely related to the prognosis of MM patients, so the ISS staging was proposed. The ISS staging is more commonly used in clinical practice due to its simplicity, but its ability to distinguish low-risk patients in the era of new treatments is limited, and it cannot guide the prognostic stratification of patients undergoing autologous stem cell transplantation. In addition, the increase in some indicators such as β2-MG may be unrelated to tumor burden and may be caused by other reasons such as impaired renal function. In 2015, based on the ISS staging, according to genetic abnormalities detected by FISH such as t(4;14), t(14;16), and del(17p), as well as the level of lactate dehydrogenase, the revised ISS staging system, namely R-ISS staging, was incorporated. R-ISS can be more effectively used for prognostic stratification of patients in different age groups and transplantation groups than ISS staging, but this study mainly came from clinical trial data, there were differences in the data analysis criteria of different clinical centers, the levels of FISH and LDH were not well calibrated, and the evaluation system did not fully consider host factors such as age, physical condition, and complications. In 2022, it was updated again to the R2-ISS system, incorporating the common chromosome 1 abnormalities in MM as prognostic indicators, but this staging system has not yet been well validated in multiple centers. The Mayo Clinic research group in 2007 based on the mSMART prognostic stratification criteria of cell molecular genetics and revised them in 2009 and 2013 (mSMART2.0 and mSMART3.0 respectively). Mainly due to the continuous addition of second-generation proteasome inhibitors (PI), monoclonal antibodies (MoAb), and immunomodulators (IMiD), the prognostic factors of MM have changed accordingly. For example, bortezomib can largely overcome the poor prognosis of the (4;14) translocation, so the t(4;14) translocation positive was adjusted from the high-risk group to the intermediate-risk group, laying the foundation for individualized treatment of MM.However, the mSMART series of risk stratification systems incorporate various technical means such as FISH, PCLI, and GEP, providing more information and guidance for prognosis and treatment. However, due to reasons such as price, technical popularity, and standardization, it has not been widely carried out, and the standardization is insufficient, with significant differences in results among different centers. It is worth noting that several genetic risk stratification systems using different genetic abnormalities have been proposed, including the International Staging System (ISS), DS staging system, IMWG, mSMART, etc., all of which have important clinical application values and do not overlap with each other.

[0005] Generally speaking, MM is a heterogeneous disease, and no single staging system can ideally cover all MM patients. Summary of the Invention

[0006] In view of this, the technical problem to be solved by the present invention is to provide a prognosis risk assessment model for multiple myeloma with a wide coverage and high accuracy, as well as its construction method and application.

[0007] The present invention provides the application of markers in constructing a prognosis risk assessment model for multiple myeloma, characterized in that the markers include genetic risk factors, age, platelet count, LDH concentration, and bone marrow plasma cell ratio, and the genetic risk factors include data on chromosome fragmentation, hyperdiploidy, chromosomal translocation, chromosomal copy number reduction, and chromosomal copy number increase, wherein:

[0008] The chromosome fragmentation means that the copy number abnormality on the same chromosome exceeds 6;

[0009] The hyperdiploidy means that the total number of chromosomes exceeds 46;

[0010] The chromosomal translocations include the translocation of chromosome 14 and chromosome 16 resulting in the fusion of MAF gene and IGH gene, the translocation of chromosome 11 and chromosome 14 resulting in the fusion of CCND1 gene and IGH gene, and the translocation of chromosome 4 and chromosome 14 resulting in the fusion of FGFR3 gene and IGH gene;

[0011] The chromosomal copy number reduction includes the deletion of the short arm of chromosome 17 region 1 band 3 or the deletion of TP53 gene, the copy number reduction in the region from the short arm of chromosome 12 region 1 band 3 sub-band 3 times sub-band to region 1 band 2 sub-band 3 times sub-band, and the copy number reduction in the region from the long arm of chromosome 22 region 1 band 2 sub-band to region 1 band 2 sub-band 3 times sub-band;

[0012] The chromosomal copy number increase includes the copy number increase in the region from the long arm of chromosome 8 region 2 band 3 sub-band 1 to region 2 band 4 sub-band 3 and the copy number increase in the region from the long arm of chromosome X region 2 band 6 sub-band 3 to region 2 band 8.

[0013] The present invention provides a prognostic risk assessment model for multiple myeloma. The prognostic risk assessment model constructs a prognostic risk assessment model for multiple myeloma based on the data of the markers in the application as input variables.

[0014] In some embodiments, the prognostic risk assessment model includes a genetic risk score model and a prognostic risk score model.

[0015] The genetic risk score model includes: Genetic risk score = 2×Z1 + Z2 + 0.5×Z3 - Z4, where

[0016] if the translocation between chromosome 14 and chromosome 16 results in the fusion of MAF gene and IGH gene being positive, then Z1 is 1; if negative, then Z1 is 0.

[0017] Z2 is the number of positive cases of the fusion of CCND1 gene and IGH gene caused by the translocation between chromosome 11 and chromosome 14, the deletion of the short arm of chromosome 17 region 1 band 3 or the deletion of TP53 gene, the reduction of the copy number in the region from sub-band 3 of sub-region 3 of band 1 of the short arm of chromosome 12 to sub-band 3 of sub-region 2 of band 1 of region 1, chromosome fragmentation, and the increase of the copy number in the region from sub-band 1 of band 3 of the long arm of chromosome 8 to sub-band 3 of band 4 of region 2. If all are negative, then Z2 is 0.

[0018] Z3 is the number of positive cases of the fusion of FGFR3 gene and IGH gene caused by the translocation between chromosome 4 and chromosome 14 and the reduction of the copy number in the region from sub-band 2 of band 1 of the long arm of chromosome 22 to sub-band 3 of band 2 of region 1. If all are negative, then Z3 is 0.

[0019] Z4 is the number of positive cases of hyperdiploidy and the increase of the copy number in the region from sub-band 3 of band 6 of the long arm of chromosome X to sub-band 3 of band 8 of region 2. If all are negative, then Z4 is 0.

[0020] The prognostic risk score model includes: Prognostic risk score = (Genetic risk score + 2) + 1.5×N1 + 0.5×N2 + 1.5×N3 + 2×N4, where

[0021] if the patient's age >= 60 years old, then N1 is 1; if the patient's age < 60 years old, then N1 is 0.

[0022] if the patient's LDH > 250 U / L, then N2 is 1; if the patient's LDH <= 250 U / L, then N2 is 0.

[0023] if the patient's platelet count < 100, then N3 is 1; if the patient's platelet count >= 100, then N3 is 0.

[0024] N4 is defined as follows: if the proportion of bone marrow plasma cells in the patient >= 60%, then N4 = 1; if the proportion of bone marrow plasma cells in the patient < 60%, then N4 = 0.

[0025] The present invention provides a method for constructing the prognostic risk assessment model, comprising the following steps:

[0026] Step 1: Collect the clinical data and biological examination data of multiple myeloma patients as the training set;

[0027] Step 2: Perform multiple imputation according to the training set described in Step 1 to obtain multiple groups of imputed datasets;

[0028] Step 3: Use the elastic net regression model to screen and obtain the best regression model, and use the best regression model to screen the training set described in Step 1 and the multiple groups of imputed datasets described in Step 2 to obtain candidate genetic-related factors, and take the intersection of the candidate genetic-related factors obtained in different groups of data to obtain the genetic risk factors;

[0029] Step 4: Perform multivariate COX regression analysis on the genetic risk factors in the multiple groups of imputed datasets to obtain the risk coefficient of each factor, and score the genetic risk factors according to the risk coefficient to further obtain the genetic risk scoring model;

[0030] Step 5: Use the genetic risk scoring model to obtain the total genetic risk score of the multiple myeloma patients;

[0031] Step 6: Perform univariate COX regression on the total genetic risk score and the training set in Step 1 in the multiple groups of imputed datasets to screen and obtain candidate variables, and perform multivariate COX regression analysis on the candidate variables. In the multivariate COX regression analysis, set the treatment plan and transplantation as stratification variables to obtain independent risk factors;

[0032] Step 7: Screen and verify the candidate variables with p-values close to 0.05 in the multivariate COX regression analysis in Step 6 to obtain other risk factors. The combination of the other risk factors, the independent risk factors and the genetic risk factors constitutes the biomarker;

[0033] Step 8: Score the biomarker in Step 7 and combine it with the genetic risk scoring model in Step 4 to obtain the prognostic risk assessment model of the multiple myeloma patients.

[0034] In some embodiments, in the step 1, the clinical data and biological examination data include the patient's gender, age, diagnosis, physical condition, date of initial diagnosis, white blood cell count, hemoglobin concentration, plasma albumin concentration, LDH concentration, serum creatinine concentration, blood calcium concentration, β2-microglobulin (β2-MG) concentration, immunoglobulin typing, serum M protein concentration, percentage of bone marrow plasma cells, percentage of biopsy plasma cells, PET-CT, bone marrow pathology, serum free light chain, chromosome number, chromosome karyotype, fluorescence in situ hybridization, chromosomal copy number variation, gene mutation, time of death, disease recurrence time, genomic complexity, treatment plan, transplantation situation, and international risk staging.

[0035] Among them, the genomic complexity is that the number of genomic copy number abnormalities >= 9.

[0036] In some specific embodiments, the training set includes 460 newly diagnosed multiple myeloma patients from the First Affiliated Hospital of Soochow University.

[0037] In some specific embodiments, in the step 6, the independent risk factors include age, platelet count, and percentage of bone marrow plasma cells.

[0038] In some specific embodiments, in the step 7, the other risk factors include LDH concentration.

[0039] The present invention provides the application of the prognostic risk assessment model or the prognostic risk assessment model constructed by the construction method in at least one of the following (1) to (3):

[0040] (1) Constructing a system for evaluating the prognostic risk of multiple myeloma;

[0041] (2) Preparing a device for evaluating the prognostic risk of multiple myeloma;

[0042] (3) Screening drugs for preventing and / or treating multiple myeloma.

[0043] The present invention provides a prognostic risk stratification system for multiple myeloma, and the prognostic risk stratification system includes:

[0044] A data acquisition module for acquiring marker data of multiple myeloma patients, and the markers include genetic risk factors, age, platelet count, LDH concentration, and percentage of bone marrow plasma cells;

[0045] A model construction module for constructing a prognostic prediction model based on the marker data;

[0046] A prognosis score determination module, configured to determine the prognosis risk score of multiple myeloma according to the input data of a multiple myeloma patient and the prognosis risk assessment model obtained by the described prognosis risk assessment model or the described construction method;

[0047] A prognosis risk assessment module, configured to determine the prognosis risk stratification result of a multiple myeloma patient according to the prognosis risk score.

[0048] In some embodiments, the prognosis risk stratification result of the multiple myeloma patient includes:

[0049] If the prognosis risk score is 0 to 2 points, output the prognosis risk stratification result corresponding to the defined low-risk group;

[0050] If the prognosis risk score is 2.5 to 4 points, output the prognosis risk stratification result corresponding to the defined medium-risk group;

[0051] If the prognosis risk score is greater than or equal to 4.5 points, output the prognosis risk stratification result corresponding to the defined high-risk group.

[0052] The present invention provides a method for prognosis risk stratification of multiple myeloma. The method is based on the described prognosis risk stratification system, and the method includes:

[0053] S1. Data collection, collecting biomarker data of a multiple myeloma patient, where the biomarkers include genetic risk factors, age, platelet count, LDH concentration, and bone marrow plasma cell ratio;

[0054] S2. Data input, inputting the feature data collected in S1 into the data collection module;

[0055] S3. Prognosis risk prediction, using the prognosis score determination module and the prognosis risk assessment module to obtain the prognosis risk stratification result of the multiple myeloma patient.

[0056] The present invention provides a device for evaluating the prognosis risk of multiple myeloma. At least one of the following ① to ③ is provided in the device:

[0057] ①. The described prognosis risk assessment model;

[0058] ②. The prognosis risk assessment model obtained by the described construction method;

[0059] ③. The described prognosis risk stratification system.

[0060] The present invention provides a computer-readable storage medium. The storage medium includes a stored program, where when the program runs, it controls the device where the storage medium is located to execute the construction method of the prognosis risk assessment model.

[0061] The present invention provides a processor for running a program, wherein when the program runs, it executes the method for constructing the prognostic risk assessment model described above.

[0062] Compared with the prior art, the prognostic risk assessment model for multiple myeloma provided by the present invention has a wider coverage, higher prediction accuracy and stratification ability. Since its sample data is derived from a large cohort of newly diagnosed multiple myeloma patients in multiple medical centers in China, compared with other similar models internationally, it is more in line with the clinical, biological and prognostic characteristics of the Chinese population and has greater advantages in the Chinese population. Description of the Drawings

[0063] Figure 1 Shows the flow chart of the present invention;

[0064] Figure 2 Shows the best screening scheme for selecting genetic risk features by elastic net;

[0065] Figure 3 Shows the genetic characteristics of NDMM patients screened by Lasso regression;

[0066] Figure 4 Shows the forest plot of multivariate COX regression. In the figure, HighGC is High genomic complexity - high genomic complexity, cytoscore is the genetic risk score, ECOG is the performance status score > 1, Age >= 60 is the age greater than or equal to 60 years old, PLT < 100 is the platelet count less than 100×10 9 / L, elevated LDH is elevated LDH, Calcium >= 2.65 is the blood calcium greater than 2.65 mM, BM_percent >= 60 is the bone marrow plasma cell ratio greater than 60%, ISS is the International Staging System for Multiple Myeloma;

[0067] Figure 5 Shows the Kaplan Meier survival curves of the new model, R-ISS and R2-ISS models in the training set and validation set;

[0068] Figure 6 Shows the ROC curves of the new model, R-ISS and R2-ISS models in the training set and validation set;

[0069] Figure 7 Shows the comparison of c-index of the new model, R-ISS and R2-ISS models in the training set and validation set;

[0070] Figure 8 Shows the Kaplan Meier survival curves of the new model in transplant patients and non-transplant patients. Detailed implementation manners

[0071] The present invention provides a prognostic risk assessment model for multiple myeloma, and a construction method and application thereof. Those skilled in the art can draw on the content of this article and appropriately improve the process parameters to implement. It should be particularly noted that all similar substitutions and modifications are obvious to those skilled in the art, and they are all regarded as included in the present invention. The methods and applications of the present invention have been described through preferred embodiments. Those skilled in the art can obviously make changes or appropriate changes and combinations to the methods and applications in this article without departing from the content, spirit and scope of the present invention to implement and apply the technology of the present invention.

[0072] The test materials used in the present invention are all ordinary commercially available products and can be purchased in the market. The present invention will be further described below in conjunction with embodiments.

[0073] The present invention provides a set of prognostic risk assessment methods and systems for risk stratification and survival probability assessment of MM patients based on some clinical and biological indexes at the initial diagnosis or before treatment of newly diagnosed MM (NDMM) patients. The data come from a large cohort of NDMM patients in multiple medical centers in China. Compared with other similar models in the world, it is more in line with the clinical, biological and prognostic characteristics of Chinese people and has greater advantages in the Chinese population.

[0074] The main inventive content and assessment steps are as follows:

[0075] Collect relevant data I of cytogenetic detection (including fluorescence in situ hybridization technology (FISH) and chromosomal microarray analysis technology (CMA)) at the initial diagnosis of MM patients. The FISH detection must include but is not limited to these three IGH rearrangement detections, including t(4;14) / FGFR3::IGH, t(11;14) / CCND1::IGH, and t(14;16) / MAF::IGH;

[0076] Collect data II such as the age, platelet concentration, LDH concentration, and bone marrow plasma cell ratio of MM patients at the initial diagnosis;

[0077] Perform cytogenetic risk scoring based on the following indicators in Data I of MM patients. If a patient has t(14;16) / MAF::IGH, record 2 points; otherwise, no score. If a patient has del(17p13) / TP53 loss, t(11;14) / CCND1::IGH, chromothripsis, Gain (8q23.1q24.3), or Loss(12p13.33p11.23), record 1 point for each; otherwise, no score. If a patient has t(4;14) / FGFR3::IGH or Loss(22q11.21q12.3), record 0.5 points for each; otherwise, no score. If a patient has hyperdiploidy (total chromosome number exceeds 46, HRD) or Gain (Xq26.3q28), record -1 point for each; otherwise, no score. Finally, add up all the above scores to obtain the genetic risk score (CytoScore).

[0078] Perform patient prognosis risk assessment based on the patient age, platelet count, LDH concentration, and bone marrow plasma cell percentage characteristics in Clinical Data II of MM patients, as well as the genetic risk score obtained in the previous step. If the age of the bone marrow plasma cell percentage of the patient is greater than or equal to 60, record 2 points; otherwise, no score. If the patient's age is greater than or equal to 60 years old and the platelet count is less than 100 (×10 9 / L), record 1.5 points; otherwise, no score. If the LDH concentration of the patient is higher than the highest limit of the measured normal reference range, record 0.5 points; otherwise, no score. After directly adding 2 points to the genetic risk score, add it to the above scoring to obtain the total prognosis risk score of the patient.

[0079] Group the patients according to the total prognosis risk score. Patients with 0 - 2 points are classified into the low-risk group, 2.5 - 4 points into the medium-risk group, and 4.5 points and above into the high-risk group.

[0080] Example 1

[0081] The present invention provides a set of prognosis risk assessment methods and systems for risk stratification and survival probability assessment of MM patients based on some clinical and biological indicators at the time of initial diagnosis or before treatment of newly diagnosed MM (NDMM) patients. The main process and technical solutions are as follows (see the flowchart in Figure 1 )

[0082] (1) Include 694 newly diagnosed multiple myeloma (NDMM) patients in the First Affiliated Hospital of Soochow University, Suzhou Hongci Hematology Hospital, and Suzhou Guangci Oncology Hospital from 2016 to 2022. Exclude 58 patients with missing prognosis time or status. Finally, 636 MM patients are included in this study.

[0083] (2) Collect various general clinical data and various clinical and biological examination data of the patients, mainly including patient name, gender, age, diagnosis, performance status, initial diagnosis date, white blood cell count, hemoglobin concentration, plasma albumin concentration, LDH concentration, serum creatinine concentration, blood calcium concentration, β2-microglobulin (β2-MG) concentration, immunoglobulin typing, serum M protein concentration, percentage of bone marrow plasma cells, percentage of biopsy plasma cells, PET-CT, bone marrow pathology, serum free light chain, chromosome number, chromosome karyotype, fluorescence in situ hybridization, chromosomal copy number variation, gene mutation, time of death, disease recurrence time, genomic complexity (defined as the number of genomic copy number abnormalities >= 9), treatment regimen, transplantation status, international prognostic staging (ISS) of the patients, etc.

[0084] (3) Perform multiple imputation (using the MICE package in R language) on the partially missing data (except for the treatment regimen) of the 636 NDMM patients included in this study to generate 5 sets of complete data.

[0085] (4) Divide the 636 NDMM patients into two groups according to different hospitals. 460 patients from the First Affiliated Hospital of Soochow University are used as the training set for establishing the prognostic assessment system, and 176 patients from other hospitals are used as the validation set for the prognostic assessment system.

[0086] (5) Use the elastic net regression model to search for the best alpha parameter in the range of 0 to 1, and find that the best model effect can be obtained when alpha is set to 1 (i.e., Lasso regression) ( Figure 2 ). Subsequently, use Lasso regression (alpha = 1) for the training set (including the 5 sets of data before and after imputation, see Figure 3Screen and evaluate the genetic risk factors of 460 NDMM patients in [ ] and their relationship with the overall survival time of the patients, in order to obtain the minimum variables that ensure the best and robust model. Finally, the variables obtained from each group of data are intersected to obtain 11 genetic factors that are stable in multiple datasets and closely related to the prognosis of the patients, including t(14;16) / MAF::IGH, t(11;14) / CCND1::IGH, t(4;14) / FGFR3::IGH, chromothripsis (defined as more than 6 copy number abnormalities on the same chromosome, chromothripsis), del(17p13) / TP53 loss, Gain (8q23.1q24.3), Loss(12p13.33p11.23), Loss (22q11.21q12.3), hyperdiploidy (defined as the total number of chromosomes exceeding 46, abbreviated as HRD), and Gain (Xq26.3q28). And the above multiple factors are further subjected to multivariate COX regression in 5 imputed datasets to obtain the risk coefficients of each factor (see Table 1). And score according to the risk coefficients of each variable. When -1 < risk coefficient < -0.5, it is counted as -1 point. When -0.5 =< risk coefficient < 0, it is counted as -0.5 point. When 0 < risk coefficient =< 0.5, it is counted as 0.5 point. When 0.5 < risk coefficient =< 1, it is counted as 1 point. When 1 < risk coefficient < 1.5, it is counted as 1.5 point. When 1.5 < risk coefficient =< 2, it is counted as 2 points. The sum of the scores of all variables is the total genetic risk score of the patient.

[0087] Table 1. Scoring Table of Genetic Risk Factors in NDMM Patients

[0088]

[0089] Note: t(14;16) / MAF::IGH: Translocation between chromosome 14 and chromosome 16 / Fusion of MAF gene and IGH gene;

[0090] t(11;14) / CCND1::IGH: Translocation between chromosome 11 and chromosome 14 / Fusion of CCND1 gene and IGH gene;

[0091] t(4;14) / FGFR3::IGH: Translocation between chromosome 4 and chromosome 14 / Fusion of FGFR3 gene and IGH gene;

[0092] Loss(17p13) / TP53: Deletion of region 1, band 3 of the short arm of chromosome 17 or deletion of TP53 gene;

[0093] Gain(8q23.1q24.3), copy number increase in the region from sub-band 1 of band 2, region 3, long arm of chromosome 8 to sub-band 3 of band 4, region 2, long arm of chromosome 8;

[0094] Loss(12p13.33p11.23), copy number decrease in the region from sub-sub-band 3 of sub-band 3, band 1, short arm of chromosome 12 to sub-sub-band 3 of sub-band 2, band 1, region 1, short arm of chromosome 12.

[0095] Loss(22q11.2q12.3), copy number decrease in the region from sub-band 2 of band 1, region 1, long arm of chromosome 22 to sub-band 3 of band 2, region 1, long arm of chromosome 22

[0096] Gain(Xq26.3q28), copy number increase in the region from sub-band 3 of band 6, region 2, long arm of chromosome X to band 8, region 2, long arm of chromosome X.

[0097] (6) The genetic risk score of the patient, together with the patient's clinical and biological factors (see step (2) above), was subjected to univariate COX regression in 5 imputed datasets to screen out factors that were significant (p < 0.05) for the patient's survival time and survival status, including treatment regimen, transplantation status, genetic risk, genomic complexity, age, performance status, platelet count, LDH concentration, calcium concentration, percentage of bone marrow plasma cells, and international ISS stage. Further, the significant factors obtained from the above univariate analysis were included in a multivariate COX regression ( Figure 4 ), with the treatment regimen and transplantation as stratification variables. The results showed that genetic risk, age >= 60, platelet count < 100, and percentage of bone marrow plasma cells >= 60% were independent prognostic factors. However, considering that the elevation of LDH (elevated LDH) was used as an evaluation index in other prognostic assessment models, it was found in this study that the elevation of LDH (p = 0.062) was also close to being statistically significant. Therefore, the evaluation performance of the model was compared in 5 imputed training datasets with and without the elevated LDH factor, and the results showed that the evaluation performance of the model could be significantly improved after adding the elevated LDH factor (the p-values of elevated LDH in the 5 datasets were 0.024, 0.018, 0.008, 0.017, and 0.012 respectively). Finally, we obtained 5 factors for prognostic risk model evaluation, namely genetic risk, age, platelet count, elevated LDH concentration, and percentage of bone marrow plasma cells.

[0098] (7) According to the hazard ratios (HRs) of the 5 factors from the above multivariate COX regression analysis (see Figure 4)(8) Calculate the restricted cubic spline (RCS) relationship between the prognostic risk score and prognosis (see

[0099] Table 2. Scoring Table for Prognostic Risk Factors in NDMM Patients

[0100]

[0101] )(8) Calculate the restricted cubic spline (RCS) relationship between the prognostic risk score and prognosis (see Figure 5 ). Define the prognostic risk score of 0 - 2 points as the low-risk group, 2.5 - 4 points as the medium-risk group, and 4.5 points and above as the high-risk group according to the analysis results (Table 3).

[0102] Table 3. Corresponding Table of Total Prognostic Risk Score and Prognostic Stratification in NDMM Patients

[0103]

[0104] )(9) Use the Kaplan-Meier survival curve and the receiver operating characteristic curve (ROC curve) to test and verify the evaluation of the survival status of patients in different risk groups by the new prognostic model in the training set and the validation set. The results show that in the training set and the validation set, the new model can accurately predict the death risk of MM patients and perform precise prognostic stratification on MM patients, and its diagnostic performance is significantly better than the R-ISS and R2-ISS staging systems (the AUC values at 1, 3, and 5 years in the training set are 0.78, 0.78, and 0.84 respectively; the AUC values at 1, 3, and 5 years in the validation set are 0.74, 0.79, and 0.83 respectively) ( Figure 6 ).

[0105] )(10) Further calculate the continuous concordance index (Time c-index) to evaluate the predictive ability of the new model and compare it with the R-ISS and R2-ISS models ( Figure 7 ). The results show that in the training set, the c-index values of the new model are higher than 0.7 in different time periods, while in the validation set, the c-index values are about 0.65 on average, indicating that the new model has a higher ability to stratify the prognosis of patients than the R-ISS and R2-ISS staging systems.

[0106] Use the Kaplan-Meier survival curve to evaluate the prognostic evaluation performance of the new model in transplanted patients and non-transplanted patientsFigure 8 ). The results show that the new model can accurately evaluate the prognosis of patients in both transplanted and non-transplanted patients.

[0107] Effect Example 1

[0108] Collect the clinical and biological indicators of Patient 1 as follows: age 68, chromosome number 48, platelet count 127×10 9 / L, LDH concentration 139.1 U / L, bone marrow plasma cell ratio 6.5%, 17p13 deletion positive, t(4;14) / FGFR3::IGH negative, t(11;14) / CCND1::IGH negative, t(14;16) / MAF::IGH negative, chromosome fragmentation negative, Gain(Xq26.3q28) positive, Gain(8q23.1q24.3) negative, Loss(12p13.33p11.23) negative, Loss(22q11.21q12.3) negative. According to the genetic risk scoring criteria in Table 1, the genetic score of this patient is -1 point. Subsequently, according to the scoring criteria in Table 2, the total prognosis risk score of this patient is 2.5 points. According to Table 3, this patient is determined to be in the intermediate-risk group.

[0109] Effect Example 2

[0110] Collect the clinical and biological information of Patient 2 as follows: age 51, chromosome number 42, platelet count 363 ×10 9 / L, LDH concentration 252 U / L, bone marrow plasma cell ratio 7%, 17p13 deletion positive, t(4;14) / FGFR3::IGH negative, t(11;14) / CCND1::IGH negative, t(14;16) / MAF::IGH positive, chromosome fragmentation negative, Gain(Xq26.3q28) negative, Gain(8q23.1q24.3) negative, Loss(12p13.33p11.23) negative, Loss(22q11.21q12.3) negative. According to the genetic risk scoring criteria in Table 1, the genetic score of this patient is 3 points. Subsequently, according to the scoring criteria in Table 2, the total prognosis risk score of this patient is 6.5 points. According to Table 3, this patient is determined to be in the high-risk group.

[0111] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art in this technical field, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. Application of the marker in constructing a prognostic risk assessment model for multiple myeloma, characterized in that: The markers include genetic risk factors, age, platelet count, LDH concentration and bone marrow plasma cell ratio, and the genetic risk factors include data on chromosome fragmentation, hyperdiploidy, chromosome translocation, chromosome copy number reduction and chromosome copy number increase, wherein: The chromosome fragmentation is that the number of abnormal copies in the same region of the same chromosome exceeds 6; The hyperdiploidy means that the total number of chromosomes exceeds 46; The chromosomal translocation includes translocation of chromosome 14 and chromosome 16 resulting in fusion of MAF gene and IGH gene, translocation of chromosome 11 and chromosome 14 resulting in fusion of CCND1 gene and IGH gene, and translocation of chromosome 4 and chromosome 14 resulting in fusion of FGFR3 gene and IGH gene; The reduced chromosome copy number includes deletion of region 1, band 3 of the short arm of chromosome 17 or deletion of the TP53 gene, reduced copy number of the region 1, band 3, subband 3, subband 3 to region 1, band 2, subband 3 of the short arm of chromosome 12, and reduced copy number of the region 1, band 1, subband 2 to region 1, band 2, subband 3 of the long arm of chromosome 22; The increased chromosome copy number includes an increase in the copy number of the region from band 1 subband in region 2 to band 4 subband in region 2 of the long arm of chromosome 8 and an increase in the copy number of the region from band 3 subband in region 2 to band 8 in region 2 of the long arm of chromosome X.

2. A prognostic risk assessment model for multiple myeloma, characterized in that: The prognostic risk assessment model is based on the data of the marker in the application of claim 1 as input variables to construct a prognostic risk assessment model for multiple myeloma.

3. The prognostic risk assessment model according to claim 2, characterized in that: The prognostic risk assessment model includes a genetic risk scoring model and a prognostic risk scoring model. The genetic risk score model includes: genetic risk score = 2×Z1+Z2+0.5×Z3-Z4, where The Z1 is 1 if the translocation of chromosome 14 and chromosome 16 results in the fusion of MAF gene and IGH gene, which is positive, and 0 if it is negative; The Z2 is: the number of positive results of translocation between chromosomes 11 and 14 leading to fusion of CCND1 gene and IGH gene, deletion of region 1, band 3 of the short arm of chromosome 17 or deletion of TP53 gene, decreased copy number of region 1, band 3, subband 3, subband 3 to region 1, band 2, subband 3, subband 1 of the short arm of chromosome 12, chromosome fragmentation, and increased copy number of region 2, band 3, subband 1 to region 2, band 4, subband 3 of the long arm of chromosome 8. If all are negative, Z2 is 0; The Z3 is the number of positive results of translocation between chromosome 4 and chromosome 14, resulting in fusion of FGFR3 gene and IGH gene and reduction of copy number from region 1, band 1, subband 2 to region 1, band 2, subband 3 of the long arm of chromosome 22. If all are negative, Z3 is 0. The Z4 is the number of positive copies of hyperdiploidy and the region from band 3 subband 2 to band 8 of band 2 of the long arm of the X chromosome. If all are negative, Z4 is 0. The prognostic risk score model includes: prognostic risk score = (genetic risk score + 2) + 1.5 × N1 + 0.5 × N2 + 1.5 × N3 + 2 × N4, where: The N1 is: if the patient's age is >= 60 years old, then N1 is 1; if the patient's age is < 60 years old, then N1 is 0; The N2 is: if the patient's LDH> 250 U / L, then N2 is 1; if the patient's LDH<= 250 U / L, then N2 is 0; The N3 is: if the patient's platelet count is <100, then N3 is 1; if the patient's platelet count is >=100, then N3 is 0; The N4 is: if the patient's bone marrow plasma cell ratio is greater than or equal to 60%, then N4 is 1; if the patient's bone marrow plasma cell ratio is less than 60%, then N4 is 0.

4. The method for constructing a prognostic risk assessment model according to claim 2 or 3, characterized in that: The steps include: Step 1, collecting clinical data and biological examination data of multiple myeloma patients as a training set; Step 2, performing multiple interpolation according to the training set described in step 1 to obtain multiple interpolation data sets; Step 3, using the elastic network regression model to screen and obtain the best regression model, using the best regression model to screen the training set described in step 1 and the multiple groups of interpolation data sets described in step 2 to obtain candidate genetic related factors, and taking the intersection of the candidate genetic related factors obtained in different groups of data to obtain the genetic risk factors; Step 4: performing a multivariate COX regression analysis on the genetic risk factors in the multiple interpolation data sets to obtain a risk coefficient for each factor, and assigning a score to the genetic risk factor according to the risk coefficient to further obtain a genetic risk scoring model; Step 5, using the genetic risk scoring model to obtain the genetic risk of the multiple myeloma patient; Step 6: Perform univariate COX regression on the genetic risk score and the training set in step 1 in the multiple imputed data sets to screen candidate variables, and perform multivariate COX regression analysis on the candidate variables, wherein treatment regimen and transplantation are set as stratification variables to obtain independent risk factors; Step 7, screening and verifying the candidate variables with p-values ​​close to 0.05 in the multivariate COX regression analysis in step 6 to obtain other risk factors, wherein the other risk factors, the independent risk factors and the genetic risk factors are combined to form the markers; Step 8: assigning scores to the markers in step 7, and combining the genetic risk equalization model in step 4 to obtain a prognostic risk assessment model for the multiple myeloma patient.

5. Use of the prognostic risk assessment model described in claim 2 or 3 or the prognostic risk assessment model constructed by the construction method described in claim 4 in at least one of the following (1) to (3): (1) Build a system for assessing the prognostic risk of multiple myeloma; (2) Preparing a device for assessing the prognostic risk of multiple myeloma; (3) Screening of drugs for the prevention and / or treatment of multiple myeloma.

6. A prognostic risk stratification system for multiple myeloma, characterized in that: The prognostic risk stratification system includes: A data collection module, used to collect marker data of multiple myeloma patients, wherein the markers include genetic risk factors, age, platelet count, LDH concentration and bone marrow plasma cell ratio; A model building module, used to build a prognosis prediction model based on the marker data; A prognostic score determination module, used to determine the prognostic risk score of multiple myeloma based on the input data of the multiple myeloma patient and the prognostic risk assessment model described in claim 2 or 3 or the prognostic risk assessment model constructed by the construction method described in claim 4; The prognostic risk assessment module is used to determine the prognostic risk stratification results of multiple myeloma patients according to the prognostic risk score.

7. The prognostic risk stratification system according to claim 6, characterized in that: The prognostic risk stratification results for multiple myeloma patients include: If the prognostic risk score is 0 to 2 points, the prognostic risk stratification result corresponding to the low-risk group is output; If the prognostic risk score is 2.5 to 4 points, the prognostic risk stratification result corresponding to the medium-risk group is output; If the prognostic risk score is greater than or equal to 4.5 points, the prognostic risk stratification result corresponding to the high-risk group is output.

8. A method for prognostic risk stratification of multiple myeloma, characterized in that: The method is based on the prognostic risk stratification system according to claim 6 or 7, and the method comprises: S1. Data collection: collecting marker data of multiple myeloma patients, including genetic risk factors, age, platelet count, LDH concentration and bone marrow plasma cell ratio; S2, data input, inputting the feature data collected in S1 into the data collection module; S3. Prognostic risk prediction: using the prognostic score determination module and the prognostic risk assessment module to obtain the prognostic risk stratification results of the multiple myeloma patient.

9. A device for assessing the prognostic risk of multiple myeloma, characterized in that The device is provided with at least one of the following ①~③: ①. The prognostic risk assessment model according to claim 2 or 3; ②. The prognostic risk assessment model constructed by the construction method described in claim 4; ③. The prognostic risk stratification system according to claim 6 or 7.

10. A computer-readable storage medium, characterized in that The storage medium includes a stored program, wherein when the program is executed, the device where the storage medium is located is controlled to execute the method for constructing a prognostic risk assessment model according to claim 4.

Images