KR102481624B1 - Personalized prognostic prediction model of intermittent treatment for chronic myeloid leukemia - Google Patents

Abstract

The present invention is an invention that can provide various techniques for establishing patient-specific mathematical modeling capable of predicting recurrence patterns after the second discontinuation of Imatinib in patients with chronic myeloid leukemia, which are different for each patient.

Description

Personalized prognostic prediction model of intermittent treatment for chronic myeloid leukemia}

The present invention relates to mathematical modeling capable of predicting the secondary recurrence pattern of cancer after discontinuation of imatinib (trade name: Gleevec), a chronic myeloid leukemia treatment agent that is different for each patient, and related technologies thereof.

Chronic Myeloid Leukemia (CML) is the main pathogenesis of excessive proliferation of leukocytes by the BCR-ABL tumor-inducing protein, which is a byproduct of the Philadelphia chromosome, caused by the transposition of chromosomes 9 and 22. In the past, anticancer chemotherapy such as hydroxylurea and immunotherapy such as interferon that inhibits the proliferation of blood cells have been used as treatments for chronic myeloid leukemia, but tyrosine kinase that specifically inhibits the activity of BCR-ABL The development of imatinib, an inhibitor, has dramatically improved treatment outcomes. Unlike existing treatments, Imatinib only selects and destroys cancer cells, so unlike conventional anticancer drugs, it does not kill normal cells, so side effects are relatively few and the treatment effect is very good.

As a method for diagnosing chronic myelogenous leukemia and measuring the response to anticancer drugs, a method for diagnosing the presence of cancer cells by detecting the BCR-ABL 1 gene using a molecular biological technique is generally used. According to research papers, about half of chronic CML patients after 5 years of imatinib treatment have undetectable minimal residual disease (a state in which the BCR-ABL 1 gene is not detected by molecular biological diagnostic techniques in the peripheral blood of cancer patients). UMRD: Undetectable minimal residual disease or MR ⁵ : 5-log reduction). Attempts to safely discontinue administration of imatinib for patients who have reached minimally undetectable residual disease after imatinib treatment due to the economic burden caused by expensive drug prices and side effects such as nausea, muscle cramps, diarrhea, and vomiting have been steadily ongoing around the world. It is being done. According to research papers, cancer relapses (first relapse) in many patients after the first discontinuation of imatinib, and after the second anticancer drug is administered to these patients, the second URMD is reached. For patients with secondary UMRD, the recurrence and pattern of leukemia after discontinuation of secondary anticancer drugs are different for each patient, with some patients experiencing a secondary recurrence within 1 year, while some patients do not have a secondary relapse for several years. Accurate prediction of the possibility of recurrence for each patient after discontinuation of administration is a very important topic in leukemia treatment.

The predictive factors for recurrence after discontinuation of imatinib based on the results of studies published at home and abroad are the duration of imatinib administration, the duration of negative BCR-ABL results after administration of imatinib, the age of the patient, the size of the spleen, and the platelet index. Factors such as sokal risk score have been reported, but these methods are predictive factors derived from statistical analysis, and have limitations in predicting the prognosis after discontinuation of imatinib, which is different for each patient. Attempts to predict the prognosis of imatinib-treated patients through mathematical modeling-based computer simulations have been made in a small number of foreign groups. A representative example is an attempt to predict the timing of leukemia recurrence with an ordinary differential equation mathematical model established based on the amount of BCR-ABL gene that decreases after imatinib administration, but predicting whether and when recurrence will occur after clinically important second imatinib discontinuation A mathematical modeling-based approach that does this has not been done yet.

In short, a method for predicting the secondary recurrence of imatinib, which is an important topic in the treatment of chronic myelogenous leukemia, has not been developed. The present invention relates to a technique for predicting the prognosis of a secondary recurrence of a patient-specific chronic myelogenous leukemia patient based on mathematical modeling.

The present inventors discovered a patient-specific mathematical model capable of predicting the recurrence pattern after the second discontinuation of imatinib in patients with chronic myeloid leukemia, which is different for each patient.

Accordingly, the present invention includes the step of measuring changes in cells, including the cell number, growth rate, differentiation rate and death rate of leukocyte hematopoietic stem cells, leukocyte progenitor cells and leukocyte differentiated cells in a sample obtained from an individual. To provide a method of providing information for predicting the prognosis of chronic myeloid leukemia recurrence after discontinuation of drug treatment.

In addition, the present invention includes the step of measuring changes in cells, including the cell number, growth rate, differentiation rate and death rate of leukocyte hematopoietic stem cells, leukocyte progenitor cells and leukocyte differentiated cells in a sample obtained from the subject. It is to provide a system for predicting the prognosis of chronic myelogenous leukemia recurrence after discontinuation of drug treatment.

The present invention developed and explained the interaction between hematopoietic stem cells, progenitor cells, and differentiated cells involved in the carcinogenesis of chronic myelogenous leukemia by developing an ordinary differential equation model. This model can predict the pattern of reduction and increase of BCR-ABL1 gene-positive cancer cells after administration and discontinuation of anticancer drugs for each patient.

Specifically, the present invention provides the step of measuring changes in cells, including the cell number, growth rate, differentiation rate, and death rate of leukocyte hematopoietic stem cells, leukocyte progenitor cells, and leukocyte differentiated cells in a sample obtained from an individual. It is to provide a method of providing information for predicting the prognosis of chronic myeloid leukemia recurrence after discontinuation of drug treatment containing the present invention.

In addition, the present invention includes the step of measuring changes in cells, including the cell number, growth rate, differentiation rate and death rate of leukocyte hematopoietic stem cells, leukocyte progenitor cells and leukocyte differentiated cells in a sample obtained from the subject. It is to provide a system for predicting the prognosis of chronic myelogenous leukemia recurrence after discontinuation of drug treatment.

The present invention provides a method or system for providing information for predicting the prognosis of chronic myelogenous leukemia recurrence, which includes calculating changes in cells according to the following equation.

<mathematical expression>

,

,

,

,

: 백혈구성 조혈모 세포,

: leukocyte hematopoietic stem cells,

: 백혈구성 전구세포,

: leukocyte progenitor cells,

: 백혈구성 분화세포,

: leukocyte differentiated cells,

: 시간,

: time,

: 백혈구성 조혈모 세포 성장 속도 ,

: leukocyte hematopoietic stem cell growth rate ,

: 백혈구성 조혈모 세포에서 전구 세포로 분화되는 속도,

: The rate of differentiation from leukocyte hematopoietic stem cells to progenitor cells,

백혈구성 조혈모 세포가 이매티닙에 의해 사멸되는 속도,

The rate at which leukocyte hematopoietic stem cells are killed by imatinib,

: 백혈구성 전구 세포가 이매티닙에 의해 사멸되는 속도,

: The rate at which leukocyte progenitor cells are killed by imatinib,

: 백혈구성 전구 세포가 백혈구성 분화 세포로 분화되는 속도,

: The rate at which leukocyte progenitor cells differentiate into leukocyte differentiated cells,

: 백혈구성 분화 세포가 자연 사멸하는 속도.

: Speed of natural death of differentiated leukocytes.

It is characterized in that the method or system provides information for predicting the prognosis of chronic myelogenous leukemia recurrence, characterized in that the recurrence is a recurrence after the second discontinuation of drug treatment.

The drug is characterized in that it is a method or system for providing information for predicting the prognosis of chronic myelogenous leukemia recurrence, characterized in that imatinib.

Characterized in that the sample is a method or system for providing information for predicting the prognosis of chronic myeloid leukemia recurrence, characterized in that the bone marrow.

In addition, the present invention is a chronic myelogenous leukemia patient's hematopoietic stem cell-progenitor cell-modeling the interaction of differentiated cells; Comparing data of BCR-ABL1/ABL1 change over time of each patient with mathematical modeling to find model parameters that minimize the error; Predicting a primary recurrence probability using a recurrence index formula; Predicting a secondary recurrence probability using a recurrence index formula; and simulating BCR-ABL1/ABL1 changes in patients with chronic myelogenous leukemia and predicting changes in the disease.

The step of finding the parameter may include calculating the number of differentiated leukocytes and the number of normal differentiated cells; Calculating an error function as a least square sum of differences between BCR-ABL1/ABL1 and BCR-ABL1/ABL1 data predicted by the model; Finding a primary parameter candidate group using a genetic algorithm; It is characterized by a modeling method for predicting the prognosis of chronic myelogenous leukemia recurrence, which includes; and finding a final parameter using a gradient descent method.

The step of estimating the primary recurrence probability includes checking the estimated parameters; It is characterized by a modeling method for predicting the prognosis of chronic myelogenous leukemia recurrence, which includes; and measuring the death rate of differentiated leukemia cells.

The step of predicting the secondary recurrence probability is a modeling method for predicting the prognosis of chronic myelogenous leukemia recurrence comprising the step of determining that the probability of recurrence within 1 year after stopping anticancer drug administration is high if the recurrence index formula <α to be characterized

The step of predicting the change of the disease may include simulating a decrease in BCR-ABL1/ABL1 of the patient during anticancer drug treatment; simulating whether to maintain BCR-ABL1/ABL1 <0.0032% after 2 or more years of anticancer drug administration; simulating the recurrence of leukemia due to the increase in BCR-ABL1/ABL1 after the first cessation of anticancer drugs; simulating a decrease in BCR-ABL1/ABL1 by administering anticancer drugs after the first recurrence; simulating maintenance of BCR-ABL1/ABL1 < 0.0032% after re-administration of anticancer drugs for 2 years or more; It is characterized by a modeling method for predicting the prognosis of chronic myelogenous leukemia recurrence that includes; and simulating changes in BCR-ABL1/ABL1 until recurrence after the second discontinuation of anticancer drugs.

The present invention is the first mathematical modeling-based computer simulation technology that can predict the discontinuation of the second treatment of a personalized targeted anticancer drug after discontinuation of treatment for patients with chronic myelogenous leukemia. It can be a technology that can predict the risk of recurrence after stopping anticancer drugs. In addition, by pre-selecting patients with a low risk of recurrence, it is expected that the cost of unnecessary administration of expensive anticancer drugs can be drastically reduced.

1 shows a schematic diagram of the technology of the present invention.
2 shows a schematic diagram of a compartmentalized mathematical model.
Figure 3 shows the change in cell number over time predicted by the mathematical model. In this patient, it was predicted that the number of progenitor cells and differentiated cells rapidly decreased when taking imatinib, but the number of hematopoietic stem cells did not change, which is related to the possibility of rapid recurrence after discontinuation of targeted anticancer drugs.
4 shows the mathematical model correction results. The change in BCR-ABL1 in this patient was corrected (grey) using the data from the first intake of imatinib to the recurrence after the first discontinuation of imatinib, and the mathematical model was corrected using this corrected model. Predicted changes in BCR-ABL1 after the 1st administration, 2nd discontinuation (blue).
Figure 5 shows an example of dividing the time of recurrence after the first cessation of the target anticancer agent according to the rate of leukocyte differentiation cell death.
Figure 6 shows a comparison of the probability of not recurring between a group (red) and a group (blue) with a small apoptosis rate of leukocyte differentiation.
Figure 7 shows an example of predicting recurrence time after the second cessation of a target anticancer drug using the ratio of leukocyte hematopoietic stem cell death rate and leukocyte progenitor cell growth rate ratio.
Figure 8 shows the distribution of indicators predicting recurrence after discontinuation of anticancer drug administration for chronic myelogenous leukemia. It can be confirmed that the distribution of the actual clinical recurrence index of the low-risk group and the high-risk group predicted by the mathematical model is significantly different based on the reference point α.
Figure 9 shows the distribution of the tendency cumulative function of the probability of non-recurrence according to the time after cessation of anticancer drug administration for the high-risk group and the low-risk group of recurrence, which are classified using the recurrence index developed in the present invention. In the case of the high-risk group (red), all patients relapsed within 1 year.

Hereinafter, preferred embodiments of the present invention will be described in detail. In addition, in the following description, many specific details such as specific components are shown, which are provided to help a more general understanding of the present invention, and it is common in the art that the present invention can be practiced without these specific details. It will be self-evident to those who have the knowledge of And, in describing the present invention, if it is determined that a detailed description of a related known function or configuration may unnecessarily obscure the subject matter of the present invention, the detailed description will be omitted.

The present invention relates to a method for providing information for predicting the prognosis of recurrence of chronic myeloid leukemia or a system for predicting the prognosis of recurrence of chronic myeloid leukemia, which includes the step of calculating cell changes according to the following equation.

<mathematical expression>

,

,

,

,

: 백혈구성 조혈모 세포,

: leukocyte hematopoietic stem cells,

: 백혈구성 전구세포,

: leukocyte progenitor cells,

: 백혈구성 분화세포,

: leukocyte differentiated cells,

: 시간,

: time,

: 백혈구성 조혈모 세포 성장 속도 ,

: leukocyte hematopoietic stem cell growth rate ,

: 백혈구성 조혈모 세포에서 전구 세포로 분화되는 속도,

: The rate of differentiation from leukocyte hematopoietic stem cells to progenitor cells,

백혈구성 조혈모 세포가 이매티닙에 의해 사멸되는 속도,

The rate at which leukocyte hematopoietic stem cells are killed by imatinib,

: 백혈구성 전구 세포가 이매티닙에 의해 사멸되는 속도,

: The rate at which leukocyte progenitor cells are killed by imatinib,

: 백혈구성 전구 세포가 백혈구성 분화 세포로 분화되는 속도,

: The rate at which leukocyte progenitor cells differentiate into leukocyte differentiated cells,

: 백혈구성 분화 세포가 자연 사멸하는 속도.

: Speed of natural death of differentiated leukocytes.

In the present invention, "recurrence of chronic myelogenous leukemia" includes cases in which clinical indicators such as hematological response, complete chromosomal response, and major genetic response are changed. Hematological responses include changes in white blood cell, red blood cell, and platelet counts. Whole chromosome reactions include detection of the Philadelphia chromosome and the like. A major gene response refers to the fact that a patient's specific mutant gene is not found even when a patient's sample is replicated 1,000 times in large quantities (PCR amplification). Preferably corresponds to the expression of BCR-ABL1. It is not limited thereto as long as it is an indicator that can be diagnosed as an additional chronic myelogenous leukemia.

In the present invention, "sample" may preferably be blood, plasma, serum, bone marrow, tissue, cell, saliva, sputum, hair, urine, etc., more preferably blood, bone marrow cells, etc., but leukocyte hematopoietic stem Any sample capable of measuring changes in cells, leukocyte progenitor cells, and differentiated leukocytes is not limited thereto.

In the present invention, "method for providing information" means the number of cells, growth rate, differentiation rate, and death rate of leukocyte hematopoietic stem cells, leukocyte progenitor cells, and leukocyte differentiated cells in a sample obtained from an individual. A method of providing information for predicting the prognosis of chronic myelogenous leukemia recurrence after drug treatment cessation including measuring changes, or information that can be confirmed by measuring changes in leukocytes is not limited thereto.

In one embodiment of the present invention, "treatment of chronic myelogenous leukemia" is not limited to any treatment method such as chemotherapy, immunotherapy, or hematopoietic stem cell transplantation. Preferably, it is by a tyrosine kinase inhibitor, more preferably, it may be a drug used as a treatment for chronic myelogenous leukemia by inhibiting the activity of tyrosine kinase, such as imatinib and nilotinib.

Hereinafter, the present invention will be described in more detail through examples. However, these examples are for illustrating the present invention, and the scope of the present invention is not limited to these examples.

< Example >

In order for a mathematical model to accurately predict a patient's secondary recurrence, it is essential to estimate parameters that can simulate changes in BCR-ABL1 gene expression for each patient. To this end, the co-inventor of this invention, Seoul St. Mary's Hospital Blood Hospital Director (Professor Kim Dong-wook), used the data of 22 patients who had been followed for 10 to 15 years. These patients had a good prognosis of imatinib treatment and were discontinued from imatinib twice. It consists of values tracking changes over time in BCR-ABL1 gene-positive cancer cells in each patient.

In the present invention, the step of modeling the interaction of hematopoietic stem cells-progenitor cells-differentiated cells of chronic myelogenous leukemia patients was preferentially performed. The effects of imatinib on leukemic hematopoietic stem cells, leukemia progenitor cells, and leukemic differentiated cells after collecting bone marrow from each patient are shown in FIG. 2 . Actual changes in each cell after administration of Imatinib are shown in FIG. 3 .

To calibrate the mathematical model, we compared the change data of BCR-ABL1/ABL1 over time in each patient with the mathematical modeling to find model parameters that minimized the error. Calculate using the number of differentiated leukocytes and the number of normal differentiated cells (FIG. 4). In order to find the mathematical model parameters that enable the mathematical model to well simulate the change in BCR-ABL1 in each patient, a nonlinear least squares function that minimizes the difference between each patient's data and the mathematical model predicted value was constructed. Data were used from the start of imatinib administration to the time of recurrence after the first discontinuation. The independent variable of this least squares function is a parameter of the model, and the function value (the error between the model and the data) changes according to the parameter value. Genetic algorithms and gradient descent were used to find parameters that minimized the function values for each patient. The final parameter values were found by constructing the primary parameter candidates using a genetic algorithm and setting them as the initial values of the gradient descent method. By substituting these parameter values into a mathematical model, the development of a patient-specific chronic myelogenous leukemia mathematical model was completed.

Using the developed patient-specific mathematical model, binary tree decision analysis was performed using all model parameters and the first recurrence time to find out what mathematical model would predict recurrence after the first discontinuation of Imatinib. It was revealed that it was possible to distinguish between a patient group with early recurrence and a patient group with delayed primary recurrence (Fig. 5-6).

In addition, a patient-specific mathematical model was used to predict the BCR-ABL1 response after imatinib administration and the increase pattern of BCR-ABL1 after the second discontinuation in patients with the first relapse, and compare them with the data, thereby reducing the error between the model prediction and the data. I made sure it wasn't too big. In addition, the mathematical model parameters and the second recurrence time were analyzed using binary tree decision analysis. revealed that It was confirmed that the higher the ratio of these two model variables (>1.4 or more), the later the second recurrence. Most of the patient groups with this small ratio value relapsed within 1 year, and especially when the death rate of differentiated cells was small (<0.13), as well as this ratio value, it was found that they recurred within 6 months (Figs. 7-9).

Through the mathematical modeling technique developed above, it is possible to predict changes in the disease after simulating changes in BCR-ABL1/ABL1 in patients with chronic myelogenous leukemia. Simulation of a decrease in BCR-ABL1/ABL1 in patients during anticancer drug treatment, simulation of maintaining BCR-ABL1/ABL1 < 0.0032% after 2 years or more of anticancer drug administration, simulation of leukemia recurrence due to an increase in BCR-ABL1/ABL1 after the first cessation of anticancer drug , Simulating the reduction of BCR-ABL1/ABL1 by anticancer drug administration after the first relapse, simulating the maintenance of BCR-ABL1/ABL1 < 0.0032% after re-administration of anticancer drugs for more than 2 years, BCR-ABL1 until recurrence after the second cessation of anticancer drugs /ABL1 can simulate changes.

Therefore, the present invention includes a mathematical model method for explaining the decrease and increase in BCR-ABL1 in chronic myelopathy patients, a method for estimating mathematical model parameters using patient data, and indicators for predicting secondary recurrence of patients using parameters can do.

Claims (15)

Measuring changes in cells, including cell numbers, growth rate, differentiation rate, and death rate, of leukocyte hematopoietic stem cells, leukocyte progenitor cells, and leukocyte differentiated cells in a sample obtained from the subject; And a method for providing information for predicting the prognosis of chronic myelogenous leukemia recurrence after treatment cessation comprising calculating changes in cells according to the following equation.

<mathematical expression>

,

,

: leukocyte hematopoietic stem cells,

: leukocyte progenitor cells,

: leukocyte differentiated cells,

: time,

: leukocyte hematopoietic stem cell growth rate ,

: The rate of differentiation from leukocyte hematopoietic stem cells to progenitor cells,

The rate at which leukocyte hematopoietic stem cells are killed by imatinib,

: The rate at which leukocyte progenitor cells are killed by imatinib,

: The rate at which leukocyte progenitor cells differentiate into leukocyte differentiated cells,

: Speed of natural death of differentiated leukocytes.
delete The method of claim 1, wherein the recurrence is a recurrence after the second discontinuation of treatment.
The method of claim 1, wherein the treatment is performed by imatinib.
The method of claim 1, wherein the sample is bone marrow.
Measuring changes in cells, including cell numbers, growth rate, differentiation rate, and death rate, of leukocyte hematopoietic stem cells, leukocyte progenitor cells, and leukocyte differentiated cells in a sample obtained from the subject; And a system for predicting the prognosis of chronic myelogenous leukemia recurrence after treatment cessation comprising calculating changes in cells according to the following equation

<mathematical expression>

,

,

: leukocyte hematopoietic stem cells,

: leukocyte progenitor cells,

: leukocyte differentiated cells,

: time,

: leukocyte hematopoietic stem cell growth rate ,

: The rate of differentiation from leukocyte hematopoietic stem cells to progenitor cells,

The rate at which leukocyte hematopoietic stem cells are killed by imatinib,

: The rate at which leukocyte progenitor cells are killed by imatinib,

: The rate at which leukocyte progenitor cells differentiate into leukocyte differentiated cells,

: Speed of natural death of differentiated leukocytes.
delete The system for predicting the prognosis of chronic myelogenous leukemia recurrence according to claim 6, wherein the recurrence is a recurrence after the second discontinuation of treatment.
The system for predicting the prognosis of chronic myelogenous leukemia recurrence according to claim 6, wherein the treatment is by imatinib.
The system for predicting the prognosis of chronic myelogenous leukemia recurrence according to claim 6, wherein the sample is bone marrow.
delete delete delete delete delete

Images