A method, apparatus, and program for determining the causative gene type of juvenile myelomonocytic leukemia (JMML) based on the distribution of monocyte morphology in the blood.

Abstract

Providing a new method for determining gene mutations in JMML. [Solution] A method for determining gene mutations in juvenile myelomonocytic leukemia (JMML), comprising the step of determining that the subject has gene mutations in the CBL group or NF1 group if the distribution of monocyte morphology in a blood sample taken from the subject is as follows: (1) if there is a lot of A and a little D, the subject is in the PTPN11_somatic group; (2) if there is a lot of B, the subject is in the PTPN11_germline group; (3) if there is a lot of C, the subject is in the KRAS group; (4) if there is little A and C and a lot of B and D, the subject is in the NRAS group; or (5) if the subject does not fall into any of the above categories (1) to (4), the subject is in the CBL group or NF1 group. A is a monocyte that has cytoplasmic vacuoles and a club-shaped nucleus; B is a monocyte that does not form a cytoplasmic vacuole and has a club-shaped nucleus; C is a monocyte that has cytoplasmic vacuoles and a round nucleus; D represents a monocyte that does not form a cytoplasmic vacuole and has a round nucleus.

Description

Technical Field

[0001] The present invention relates to a method, apparatus, and program for determining a causative gene type of juvenile myelomonocytic leukemia (JMML) based on the distribution of monocytic forms in blood.

Background Art

[0002] Juvenile myelomonocytic leukemia (JMML) is a type of childhood leukemia, accounting for 1-2% of all childhood leukemias. In JMML, morphological features in the peripheral blood smear include an increase in abnormal monocytes, an increase in myeloblast progenitor cells and erythroblast progenitor cells, etc. (Non-Patent Document 1). Also, it is known that most cases are associated with gene abnormalities, and there are multiple types of such gene mutations. Depending on the presence or type of gene mutation, there are cases where the symptoms are severe and hematopoietic stem cell transplantation is required, or cases where the symptoms are relatively mild and follow-up observation is carried out by drug administration. Therefore, the diagnosis of gene mutation is necessary for determining the treatment policy. So far, the types of gene mutations in JMML have been determined by methods such as the Sanger sequencing method, whole exome analysis, and whole genome analysis. However, there is a problem that such methods are limited to implementation in specific facilities and are not tests that can be easily performed in all facilities. Especially in regional areas, due to the poor accessibility, there was a risk of delay in determining the treatment policy. A method for calculating polygenic scores (PGSs) from blood cell morphology has been reported (Non-Patent Document 2). On the other hand, there is no method for calculating PGSs for monocytes, and there has been no report so far on the relationship between the morphological characteristics of abnormal monocytes and gene mutations.

Prior Art Documents

Non-Patent Documents

[0003]

Non-Patent Document 1

[0004] The present invention aims to provide a novel method for determining gene mutations in JMML. Furthermore, the present invention aims to provide a JMML gene mutation determination program for causing a computer to execute the aforementioned method, and a JMML gene mutation determination device equipped with the program. [Means for solving the problem]

[0005] The present inventors, after diligently investigating the relationship between the morphological characteristics of abnormal monocytes in JMML and gene mutations, discovered that by examining the morphology of the monocyte nucleus and the presence or absence of vacuole formation in the cytoplasm of monocytes for each JMML gene mutation, it is possible to determine the JMML gene mutation from peripheral blood smears, thus completing the present invention. That is, the present invention includes the following embodiments.

[0006] Section 1. A method for determining gene mutations in juvenile myelomonocytic leukemia (JMML), The distribution of monocyte morphology in blood samples taken from the subjects was: (1) If A is abundant and D is scarce, then it belongs to the PTPN11_somatic group; (2) If B is abundant, the PTPN11_germline group; (3) If C is abundant, the KRAS group; (4) If A and C are few and B and D are many, then the NRAS group; or (5) If the case does not fall into any of the above categories (1) to (4), the CBL group or NF1 group The process includes determining whether the subject possesses each of the following gene mutations: The aforementioned A is a monocyte having a cytoplasmic vacuole and a club-shaped nucleus; The aforementioned B is a monocyte that does not form a cytoplasmic vacuole and has a club-shaped nucleus; The aforementioned C is a monocyte having a cytoplasmic vacuole and a round nucleus; and The aforementioned D is a monocyte that does not form a cytoplasmic vacuole and has a round nucleus. A method to demonstrate each of these. Section 2. If (1) above has many (1')A and C and almost no D, then it is the PTPN11_somatic group; If (2) above contains a large amount of (2')A and B, and almost no D, then it is the PTPN11_germline group; If (3) above has a large amount of (3')C and almost no B and D, then it is the KRAS group; If (4) above is the case where (4')A and C are almost absent and B and D are abundant, then it is the NRAS group; or If (5) above does not fall into any of (1') to (4'), then the CBL group or NF1 group. The method according to item 1, which involves determining whether the subject possesses each of the gene mutations. Section 3. The method according to item 1 or 2, wherein, in (1) above, if the radar chart plotted with the normalized ratios of A in the positive y-axis direction, B in the negative y-axis direction, C in the negative x-axis direction, and D in the positive x-axis direction approximates a radar chart where A=1.00, B=0.13, C=0.82, and D=0.00, it is determined to be in the PTPN11_somatic group. Section 4. The method according to any one of items 1 to 3, wherein, in (2) above, if the radar chart plotted with the normalized ratios of A in the positive y-axis direction, B in the negative y-axis direction, C in the negative x-axis direction, and D in the positive x-axis direction approximates a radar chart where A=0.85, B=1.00, C=0.38, and D=0.06, it is determined to be part of the PTPN11_germline group. Section 5. The method according to any one of items 1 to 4, wherein, in (3) above, if the radar chart plotted with the normalized ratios of A in the positive y-axis direction, B in the negative y-axis direction, C in the negative x-axis direction, and D in the positive x-axis direction approximates a radar chart where A=0.50, B=0.00, C=1.00, and D=0.02, it is determined to be a KRAS group. Section 6. The method according to any one of items 1 to 5, wherein, in (4) above, if the radar chart plotted with the normalized ratios of A in the positive y-axis direction, B in the negative y-axis direction, C in the negative x-axis direction, and D in the positive x-axis direction approximates a radar chart where A=0.00, B=0.84, C=0.00, and D=1.00, it is determined to be an NRAS group. Section 7. The method according to any one of terms 1 to 6, wherein a monocyte having a club-shaped nucleus means a monocyte whose maximum distance calculated by Convex hull is greater than the cutoff value calculated from the ROC curve used to create the model. Section 8. The method according to any one of items 1 to 6, wherein monocytes with cytoplasmic vacuole formation are defined as monocytes in which the number of vacuoles identified by HoughCircles in OpenCV is greater than the cutoff value calculated from the ROC curve used for model creation. Section 9. The method according to any one of items 1 to 8, wherein the distribution of the monocyte morphology is obtained using at least 100 monocytes. Section 10. A device for determining gene mutations in juvenile myelomonocytic leukemia (JMML), The distribution of monocyte morphology in blood samples taken from the subjects was: (1) When there are many A and few D, the PTPN11_somatic group; (2) When there are many B, the PTPN11_germline group; (3) When there are many C, the KRAS group; (4) When A and C are few and B and D are many, the NRAS group; or (5) When not classified into any of (1) to (4) above, the CBL group or the NF1 group and includes a determination unit that determines that the subject has a gene mutation of: The A is a monocyte with cytoplasmic vacuolization and a nucleus in the shape of a club; The B is a monocyte without cytoplasmic vacuolization and a nucleus in the shape of a club; The C is a monocyte with cytoplasmic vacuolization and a round nucleus; and The D is a monocyte without cytoplasmic vacuolization and a round nucleus respectively indicating, an apparatus. Item 11. A program for operating a computer as the determination unit of the apparatus according to Item 10.

Effect of the Invention

[0007] The present invention provides a method for determining gene mutations of JMML from the distribution of monocyte morphology.

Brief Description of the Drawings

[0008] [Figure 1] It is a diagram showing an example of normal monocytes and abnormal monocytes. [Figure 2] It is a diagram showing the classification of abnormal monocytes. [Figure 3A] It is a diagram showing representative images of abnormal monocytes of types A to D and the ratio of abnormal monocytes classified into A to D for each gene mutation. [Figure 3B] It is a diagram showing the ratio of abnormal monocytes classified into A to D for each gene mutation. Similar to Fig. 3A, it is shown in the order of type A, B, C, and D from the leftmost column. [Figure 4]This is a radar chart showing the normalized proportion of abnormal monocytes for each gene mutation. [Figure 5] This figure schematically shows the general configuration of a determination system according to one embodiment. [Figure 6] This is a block diagram illustrating the function of a determination device according to one embodiment. [Figure 7] This is a flowchart illustrating the data processing procedure performed by a determination device according to one embodiment. [Modes for carrying out the invention]

[0009] 1. Judgment method The present invention relates to a method for determining gene mutations in juvenile myelomonocytic leukemia (JMML), The distribution of the monocyte morphology is, (1) If A is abundant and D is scarce, then it belongs to the PTPN11_somatic group; (2) If B is abundant, the PTPN11_germline group; (3) If C is abundant, the KRAS group; (4) If A and C are few and B and D are many, then the NRAS group; or (5) If the person does not fall into any of the above categories (1) to (4), the CBL group or NF1 group; The process includes determining whether the gene has a mutation, A is a monocyte with cytoplasmic vacuole formation and a club-shaped nucleus. B is a monocyte with a club-shaped nucleus that does not form a cytoplasmic vacuole. C is a monocyte with a round nucleus, accompanied by the formation of a cytoplasmic vacuole. D is a monocyte with a round nucleus that does not form a cytoplasmic vacuole. This relates to a method for determining gene mutations. The following explains this method.

[0010] The gene mutations that can be determined by the method of the present invention include the PTPN11_somatic, PTPN11_germline, NRAS (sometimes referred to as NRAS_mutated), KRAS, CBL, and NF1 genes. Individuals with abnormalities in these genes are referred to as the PTPN11_somatic group, PTPN11_germline group, NRAS group, KRAS group, CBL group, and NF1 group, respectively. These six gene mutations are representative of the gene abnormalities found in JMML, accounting for 80-90% of the total. Therefore, the present invention makes it possible to determine almost all gene abnormalities in JMML, making it very useful.

[0011] The blood sample according to the present invention is blood collected from a JMML patient, preferably peripheral blood. The amount of peripheral blood should be between 100 μL and 1000 μL, with approximately 200 μL being commonly used.

[0012] The monocytes according to this invention are abnormal monocytes found in JMML patients. Abnormal monocytes typically have a different nuclear morphology from normal monocytes, and some form vacuoles in the cytoplasm (Figure 1).

[0013] 1.1 Preparation of blood samples This invention determines gene mutations by visualizing the morphology of abnormal monocytes and then obtaining the distribution of their morphology. When visualizing the morphology of monocytes, a blood sample is prepared on a suitable observation device. The observation device is not particularly limited as long as it can observe monocytes in the blood, and examples include glass slides, microscope slides, and microfluidic devices. The nuclei and cytoplasm of the monocytes on the observation device may be stained with an appropriate dye as needed. The blood sample may be fixed on the observation device and then stained by May-Giemsa staining, or a blood sample that has been previously stained with an appropriate dye may be applied to the observation device. The blood sample fixed on the observation device does not need to be stained. In this invention, a glass slide coated with peripheral blood is called a blood smear.

[0014] 1.2 Acquisition of blood smear images A blood smear is prepared using a blood sample, stained as necessary, and an image of the blood smear (hereinafter sometimes simply referred to as "image") is obtained. For obtaining the smear image, a microscope or other device capable of taking photographs (such as a camera) is used. Next, abnormal monocytes are identified from the obtained image. Identification of abnormal monocytes may be performed by human visual inspection or computer-based image processing using the acquired blood smear image data. Alternatively, abnormal monocytes may be identified by human visual inspection using a microscope before obtaining the blood smear image.

[0015] 1.3 Morphology of Monocytes Monocyte morphology refers to the shape of the monocyte nucleus and the presence or absence of vacuoles in the cytoplasm of the monocyte, and can be classified into four types, A to D, as described below.

[0016] In this invention, the shape of the monocyte nucleus is determined by whether or not it has a club-shaped nucleus or a round nucleus. As shown in Figure 1, normal monocytes have an elliptical nucleus with a constricted portion (Figure 1A), while abnormal monocytes may have a club-shaped nucleus with a large internal depression, elongated and curved (Figure 1B; hereinafter referred to as a club-shaped nucleus in this specification), or a rounded nucleus with multiple constrictions (Figure 1C; hereinafter referred to as a round nucleus). A club-shaped nucleus is defined as a nucleus having one depression within the nucleus, with a larger area of ​​depression compared to a normal monocyte. The depression is a depression in the cell nucleus extending from the outer periphery to the central region, and does not mean that the outer periphery of the nucleus is wavy. A round nucleus is a rounded nucleus that does not have a depression within the nucleus. Furthermore, a round nucleus includes those with a wavy outer periphery. To determine whether the nucleus of an abnormal monocyte is club-shaped or round, this can be done by visual inspection or by running software on a computer. When using a computer, the morphology of the monocyte nucleus can be determined by following the steps 1) to 3) below.

[0017] Step 1) Create an image of abnormal monocytes extracted from the blood sample obtained in 1.2. Abnormal monocytes may be extracted by human visual inspection or by computer-aided image processing. Examples of software used for image processing include MATLAB®.

[0018] Step 2) Increase the contrast of the periphery to obtain a grayscale or black and white binarized image in which the nucleus, cytoplasm, and vacuoles, which are the image features of the present invention, can be identified. The software used in Step 2) is not limited, but ilastik is preferred, for example.

[0019] Step 3) Extract only the nuclear region and evaluate the size of the nuclear depression (roughness). The software used in Step 3) is not limited as long as it can determine the roughness, but it is preferable to use Convex Hull, which can measure the maximum distance from the outer edge of the nucleus to the most depressed region. In this invention, monocytes whose maximum distance is greater than the cutoff value calculated from the ROC curve during model creation are judged to have a club-shaped nucleus, and monocytes whose maximum distance is less than the cutoff value are judged to have a round-shaped nucleus.

[0020] The presence or absence of vacuoles in the cytoplasm of monocytes can be determined by visual inspection by a human, or by using software on a computer. When using a computer, the presence or absence of vacuoles in the cytoplasm can be determined by following the steps 1), 2), and / or 3) described above, in addition to step 4) below.

[0021] Step 4) Extract vacuoles from the image in Step 2) above to identify the presence or absence of vacuoles in the cytoplasm. The software used in Step 4) is not limited as long as it can extract circles, but for example, HoughCircles in OpenCV (Yuen, H., Princen, J., Illingworth, J., Kittler, J.: Comparative study of hough transform methods for circle finding. Image and Vision Computing 8(1), 71-77 (1990)) is preferred. In this specification, nuclei having vacuoles are referred to as having cytoplasmic vacuole formation. It is preferable that monocytes having cytoplasmic vacuole formation have more cytoplasmic vacuoles than the cutoff value calculated from the ROC curve used when creating the model.

[0022] Once the monocyte's nuclear shape and the presence or absence of cytoplasmic vacuoles are determined as described above, each abnormal monocyte is classified into types A to D below (Figure 2).

[0023] A: A monocyte that has cytoplasmic vacuoles and a club-shaped nucleus; B: Monocytes that do not form cytoplasmic vacuoles and have a club-shaped nucleus; C: Monocytes that have cytoplasmic vacuoles and a round nucleus; and D: A monocyte that does not form a cytoplasmic vacuole and has a round nucleus.

[0024] The types A to D obtained in this way are called monocyte morphologies.

[0025] 1.4 Distribution of monocyte morphology The distribution of monocyte morphology is calculated from the obtained monocyte morphology. For each patient, the proportion of abnormal monocytes classified as A to D above to the total number of monocytes is plotted with A on the positive y-axis, B on the negative y-axis, C on the negative x-axis, and D on the positive x-axis, thereby obtaining a monocyte morphology distribution (radar chart) as shown in Figures 4(i) to (l). In this invention, it is preferable to use at least 100 of the abnormal monocytes to calculate the monocyte morphology distribution. In this specification, the monocyte morphology distribution is referred to as a radar chart.

[0026] 1.5 Determination of Genetic Abnormalities Finally, the distribution of monocyte morphology obtained above is: (1) If A is abundant and D is scarce, then it belongs to the PTPN11_somatic group; (2) If B is abundant, the PTPN11_germline group; (3) If C is abundant, the KRAS group; (4) If A and C are few and B and D are many, then the NRAS group; or (5) If the case does not fall into any of the above categories (1) to (4), the CBL group or NF1 group By making this determination, it is possible to identify genetic abnormalities from the distribution of monocyte morphology. Preferably, (1) If A and C are abundant and D is almost nonexistent, then it is the PTPN11_somatic group; (2) If A and B are abundant and D is almost absent, then the PTPN11_germline group; (3) If C is abundant and B and D are almost absent, then it is the KRAS group; (4) If A and C are almost absent and B and D are abundant, then the NRAS group; or (5) If the case does not fall into any of the above categories (1) to (4), the CBL group or NF1 group That is the case.

[0027] More preferably, (1) to (4) above can be defined using the proportion of the area occupied by each quadrant in a radar chart plotting the proportions of A to D as follows. In this specification, in a radar chart plotting the proportions of A, B, C, and D, the intersection of the x and y axes of the radar chart is defined as zero point O, and the upper right (OAD) region enclosed by the x and y axes is called the first quadrant, the upper left (OAC) is called the second quadrant, the lower left (OBC) is called the third quadrant, and the lower right (OBD) is called the fourth quadrant. For example, the area in the first quadrant can be obtained by multiplying the proportions of A and D in the total monsphere and dividing by 2. The areas in the other quadrants can be obtained in the same way. The proportion of the area is, In case (1), the area occupied by the second quadrant of the radar chart is the largest, followed by the area occupied by the third quadrant, with almost no area occupied by the second and fourth quadrants; (2) In this case, the proportions occupied by the second and third quadrants of the radar chart are equal, and are higher than the proportions occupied by the first and fourth quadrants; (3) In this case, the proportion occupied by the second quadrant of the radar chart is the largest, and the area occupied by the other quadrants is small; or In case (4), the area occupied in the fourth quadrant is the largest, and the area occupied in the other quadrants is almost negligible. It is preferable to select such that the following conditions are met. More preferably, when using the above area, with respect to the entire radar chart, (1) The area occupied by the first quadrant accounts for 0-10% of the total area; The second quadrant accounts for 80-100% of the area; The proportion of the area occupied by the third quadrant is 5-15%; and If the proportion of the area occupied by the fourth quadrant is 0-10%, then it belongs to the PTPN11_somatic group; (2) The proportion of the area occupied by the first quadrant is 0-10%; The second quadrant accounts for 25-45% of the total area; The proportion of the area occupied by the third quadrant is 40-60%; and If the proportion of the area occupied by the fourth quadrant is 0-10%, then it belongs to the PTPN11_germline group; (3) The proportion of the area occupied by the first quadrant is 0-10%; The second quadrant accounts for 90-100% of the area; The proportion of the area occupied by the third quadrant is 0-10%; and If the proportion of the area occupied by the fourth quadrant is 0-10%, then it is the KRAS group; (4) The proportion of the area occupied by the first quadrant is 0-10%; The proportion of the area occupied by the second quadrant is 0-10%. The proportion of the area occupied by the third quadrant is 0-10%; and If the proportion of the area occupied by the fourth quadrant is 90-100%, then it is the NRAS group; or (5) If the case does not fall into any of the above categories (1) to (4), the CBL group or NF1 group By making this determination, it is preferable to assess genetic abnormalities from the distribution of monocyte morphology. The method of the present invention can determine or predict genetic abnormalities from peripheral blood smears derived from JMML patients without undergoing genetic diagnosis.

[0028] Most preferably, (1), (2), (3), and (4) above are determined by referring to the radar chart shown in Figure 4. 100 abnormal monocytes are counted from the new case for which a genetic abnormality is to be estimated, and the proportions of A, B, C, and D are calculated. After calculation, a radar chart is created, and if there is one that approximates the radar chart shown in Figure 4, it is determined (predicted) that the case has the corresponding genetic mutation. If there is no approximation to any radar chart, it is predicted to be either the CBL group or the NF1 group. In this specification, when we say that radar charts are similar, it means that the shapes of the radar charts are similar. Specifically, it means that when the radar chart is divided into quadrants from the first to the fourth quadrant, the proportion of the area occupied by each quadrant is close. A close proportion of area means that each area ratio falls within the range of the more preferred embodiment described above, but is not necessarily limited to the above range. For example, the radar chart shown in Figure 4(i) has A=1.00, B=0.13, C=0.82, and D=0.00, where the area occupied by the first quadrant is 0.0%, the area occupied by the second quadrant is 88.5%, the area occupied by the third quadrant is 11.5%, and the area occupied by the fourth quadrant is 0.0% of the total area of ​​the radar chart. These percentages are approximate values, and values ​​within ±10%, preferably ±5%, and more preferably ±3% of each percentage are considered to approximate the region in question.Similarly, in the radar chart shown in Figure 4(j), where A=0.85, B=1.00, C=0.38, and D=0.06, the area occupied by the first quadrant is 6.3%, the area occupied by the second quadrant is 39.7%, the area occupied by the third quadrant is 46.7%, and the area occupied by the fourth quadrant is 7.4%. In the radar chart shown in Figure 4(k), where A=0.50, B=0.00, C=1.00, and D=0.02, the area occupied by the first quadrant is 2.0%, the area occupied by the second quadrant is 98.0%, and the third quadrant is 7.4%. The area occupied by the first quadrant is 0.0%, and the area occupied by the fourth quadrant is 0.0%. In the radar chart shown in Figure 4(l), relative to the entire radar chart with A=0.00, B=0.84, C=0.00, and D=1.00, the area occupied by the first quadrant is 0.0%, the area occupied by the second quadrant is 0.0%, the area occupied by the third quadrant is 0.0%, and the area occupied by the fourth quadrant is 100.0%. However, these percentages are approximate values, and if the value is within ±10%, preferably ±5%, and more preferably ±3% of each percentage, it is judged to be an approximation of the region.

[0029] The method according to the present invention does not require whole-genome sequencing using the Sanger method and can determine gene mutations from peripheral blood samples alone, making it usable regardless of location. Even in areas where conventional gene mutation determination is difficult, it is possible to predict treatment strategies and initiate treatment early. It is also expected to be useful in predicting disease prognosis. The PTPN11_somatic group, PTPN11_germline group, NRAS group, and KRAS group are relatively severe, and it is important to make decisions such as transplantation early, so this method is particularly useful in determining gene mutations in these groups.

[0030] 2. Judgment device The determination device of the present invention is a device that processes data in accordance with the determination method of the present invention. Unless otherwise specified, the processing performed by the determination device of the present invention is the same as that of the determination method of the present invention, so redundant explanations will be omitted.

[0031] 2.1 Device configuration Figure 5 is a schematic diagram showing the general configuration of a determination system according to one embodiment.

[0032] A determination system 100 according to one embodiment comprises a determination device 1 and a server device 90. The determination device 1 and the server device 90 are connected to each other via a network 9 so as to be able to communicate data. The server device 90 records various data about a subject that has been measured or acquired at medical institutions such as clinics, general hospitals, and public health centers, or at research institutions such as companies and universities. These various data about the subject are recorded as the subject's medical information or medical record, for example, in the form of a medical record (chart). The server device 90 may be installed inside the above-mentioned medical institutions or research institutions, or it may be cloud-based and installed outside the above-mentioned medical institutions or research institutions.

[0033] In this embodiment, the determination device 1 acquires various data about the subject from the server device 90 to obtain the distribution of the subject's monocyte morphology. As various data about the subject, the determination device 1 acquires blood smear image data 21, which will be described later, from the server device 90. Instead of blood smear image data, images of abnormal monocytes may be acquired from the server device 90.

[0034] Figure 6 is a block diagram illustrating the function of a determination device according to one embodiment.

[0035] A determination device 1 according to one embodiment of the present invention comprises a data processing unit 10, an auxiliary storage device 20, an input unit 31, a display unit 32, and a communication interface unit (communication I / F unit) 33. The determination device 1 can be configured using, for example, a general-purpose computer such as a notebook computer, or, for example, a tablet terminal or smartphone (hereinafter referred to as "tablet terminal, etc.").

[0036] In this embodiment, the determination device 1 comprises, as hardware components, an auxiliary storage device 20, an input unit 31, a display unit 32, and a communication I / F unit 33. Although not shown, the determination device 1 further comprises, as hardware components, a processor such as a CPU that performs data processing, and memory used by the processor as a work area for data processing.

[0037] The auxiliary storage device 20 is a non-volatile storage device that stores the operating system (OS), various control programs, and data generated by the programs, and is composed of, for example, flash memory, eMMC (embedded Multi Media Card), SSD (Solid State Drive), etc. In this embodiment, the auxiliary storage device 20 stores blood smear image data 21, abnormal monocyte image extraction parameters 22, abnormal monocyte classification parameters 23, and judgment program 29.

[0038] The blood smear image data 21 contains information about the subject's blood smear. In this embodiment, the blood smear image data 21 includes information about images of abnormal monocytes extracted from the blood smear. The abnormal monocyte image extraction parameters 22 contain information about image analysis parameters used when extracting abnormal monocytes from the blood smear image. The abnormal monocyte classification parameters 23 contain information about image analysis parameters used for classifying monocyte morphology. The image analysis parameters are predetermined.

[0039] The judgment program 29 is a computer program for realizing the parts 11 to 18 within the data processing unit 10, which will be described later, and is a software-based functional block. These functional blocks are realized by installing the judgment program 29 into the auxiliary storage device 20 or memory of the judgment device 1, and by the processor executing the judgment program 29. The judgment program 29 may also be installed into the judgment device 1 via a network 9 such as the Internet, which is connected by the communication I / F unit 33. Alternatively, the judgment program 29 may be installed into the judgment device 1 by having the judgment device 1 read a computer-readable, non-temporary, tangible recording medium, such as a memory card, on which the judgment program 29 is recorded. The judgment program 29 can be, for example, an application for a tablet terminal. The judgment program 29 can also be implemented using spreadsheet software such as Microsoft Excel® or Google® Sheets.

[0040] The input unit 31 can be configured with, for example, a mouse or keyboard, and the display unit 32 can be configured with, for example, a liquid crystal display or an organic EL display. The input unit 31 and the display unit 32 can also be integrated as a touch panel.

[0041] The communication interface unit 33 transmits and receives data with external devices such as a server device 90 via a wired or wireless network 9. The communication interface unit 33 may use various wireless or wired connections such as Bluetooth®, Wi-Fi®, and Ethernet®.

[0042] In this embodiment, the determination device 1 includes a data processing unit 10 as part of its software configuration. The data processing unit 10 is a functional block realized by the processor executing the determination program 29.

[0043] The acquisition unit 11 acquires image data 21 of the subject's blood smear. In this embodiment, the acquisition unit 11 acquires the image data 21 of the subject's blood smear from the server device 90 via the network 9. In other embodiments, the acquisition unit 11 acquires the image data 21 of the subject's blood smear via the input unit 31 based on input from an operator.

[0044] The extraction unit 12 extracts abnormal monocyte images from the subject's blood smear image data 21 based on the acquired abnormal monocyte image extraction parameters 22.

[0045] The classification unit 13 consists of a nuclear morphology classification unit 131, a cytoplasmic vacuole classification unit 132, and a type classification unit 133. The nuclear morphology classification unit 131 classifies the nuclear shape of extracted abnormal monocyte images into club-shaped or round-shaped based on abnormal monocyte classification parameters 23. The cytoplasmic vacuole classification unit 132 classifies the presence or absence of cytoplasmic vacuoles based on abnormal monocyte classification parameters 23. The type classification unit 133 classifies each abnormal monocyte of the subject into A to D based on the classified nuclear shape and the presence or absence of cytoplasmic vacuoles; A: monocyte with cytoplasmic vacuole formation and a club-shaped nucleus, B: monocyte without cytoplasmic vacuole formation and a club-shaped nucleus, C: monocyte with cytoplasmic vacuole formation and a round-shaped nucleus, D: monocyte without cytoplasmic vacuole formation and a round-shaped nucleus. In classifying abnormal monocytes, a machine learning-based classification model may be used to classify monocyte morphologies into categories A to D.

[0046] The distribution calculation unit 14 calculates the number of each type A to D of abnormal monocytes obtained and plots the ratio of each type of abnormal monocyte to the total number of abnormal monocytes for each subject, thereby calculating the distribution of monocyte morphology.

[0047] The determination unit 15 determines that the distribution of the obtained monocyte morphology is (1) If A is abundant and D is scarce, then it belongs to the PTPN11_somatic group; (2) If B is abundant, the PTPN11_germline group; (3) If C is abundant, the KRAS group; (4) If A and C are few and B and D are many, then the NRAS group; or (5) If the case does not fall into any of the above categories (1) to (4), the CBL group or NF1 group It is determined that the individual has the gene mutation.

[0048] The output unit 16 outputs the calculated gene mutation as the determination result. In this embodiment, the output unit 16 displays the determination result on, for example, the display unit 32. In other embodiments, the output unit 16 transmits the determination result to, for example, a server device 90. The server device 90 records the gene mutation of the subject output from the determination device 1, for example, in the form of a medical record (chart), associating it with the subject's medical information or other medical information.

[0049] This will allow healthcare professionals involved in the treatment and health guidance of subjects at medical institutions such as clinics, general hospitals, and public health centers, as well as research institutions, to refer to gene mutations determined from blood smears to make more appropriate diagnoses and treatment decisions.

[0050] 2.2 Processing Procedure Figure 7 is a flowchart illustrating the data processing procedure performed by a determination device according to one embodiment.

[0051] In step S1, the acquisition unit 11 acquires image data 21 of the subject's blood smear. In step S2, the extraction unit 12 extracts abnormal monocyte images using abnormal monocyte image extraction parameters. In step S3, the classification unit 13 classifies the morphology of abnormal monocytes into types A to D based on the shape of the nucleus and the presence or absence of cytoplasmic vacuoles. In step S4, the distribution calculation unit 14 calculates the distribution of monocyte morphology for each subject.

[0052] In step S5, the determination unit 15 determines gene mutations based on the calculated distribution of monocyte morphology.

[0053] In step S6, the output unit 16 outputs the determined gene mutation as the determination result. The determination result is displayed, for example, on the display unit 32.

[0054] As described above, according to the determination method and determination device of one embodiment, gene mutations can be determined or predicted from peripheral blood smear images of a subject without relying on expensive equipment or genome analysis, which has limited access. As a result, medical professionals involved in the treatment and health guidance of subjects at medical institutions and research institutions such as clinics, general hospitals, and public health centers can refer to the gene mutations determined from peripheral blood smears to make more appropriate diagnoses and decisions regarding treatment plans.

[0055] 3. Other forms Although the present invention has been described above with reference to specific embodiments, the present invention is not limited to the embodiments described above.

[0056] In the above embodiment, the determination device 1 is implemented as a single unit, but the determination device 1 does not need to be a single unit; the processor, memory, auxiliary storage device 20, etc., may be located separately and connected to each other via a network. Similarly, the input unit 31 and the display unit 32 do not necessarily need to be located in the same place; they may be located separately and connected to each other via a network for communication.

[0057] In the above embodiment, each functional block 11 to 18 constituting the data processing unit 10 is implemented by software, but each of these functional blocks 11 to 18 may be partially or entirely implemented as hardware. The processing of each functional block 11 to 18 constituting the data processing unit 10 does not need to be processed by a single processor, but may be distributed and processed by multiple processors. The functions of the data processing unit 10 and the data items in the auxiliary storage device 20 may be partially or entirely cloudified on another server device (not shown) connected via the communication I / F unit 33.

[0058] In one embodiment of the present invention, images of abnormal monocytes may be generated by processing images of a blood smear with a microscope or a computing device connected to a server device 90, and these images of abnormal monocytes may be stored in the server device 90. The classification unit may then classify the abnormal monocytes, and the distribution calculated by the distribution calculation unit 14 may be used to determine the genetic abnormality. In this case, an extraction unit that extracts images of abnormal monocytes using abnormal monocyte image extraction parameters is not necessary.

[0059] In other embodiments of the present invention, at least one of the steps selected from the group consisting of the acquisition unit 11, extraction unit 12, classification unit 13, distribution calculation unit 14, and determination unit 15 may be performed by human operation.

[0060] In other embodiments of the present invention, the determination device 1 may be connected to a device that acquires image data of a blood smear in a communicative manner. The measuring device that acquires images of abnormal monocytes may itself have a function to determine genetic abnormalities from the acquired images. [Examples]

[0061] The present invention is further illustrated by the following embodiments, which should not be construed as further limitations.

[0062] The analysis included 92 patients from JMML cases who were registered in a central diagnostic registry, had residual smears, identified gene mutations, and underwent methylation analysis. The breakdown of patients with each gene abnormality is shown in Table 1.

[0063] [Table 1]

[0064] 100 abnormal monocytes were extracted from each patient's peripheral blood smear, for a total of 9200 abnormal monocytes. These were classified into types A through D as defined above. The overall breakdown of the four defined abnormal monocyte classifications was as follows: Type A: 23.6%, Type B: 19.4%, Type C: 30.7%, and Type D: 26.3%. Next, the proportion of A through D was calculated for each gene mutation, and the relationship between gene abnormalities and monocyte morphology was examined. The results of comparing each type (A through D) for patients with each gene mutation are shown in Figures 3A and 3B.

[0065] Figure 3A(e) shows that the proportion of abnormal type A monocytes was significantly higher in patients with the PTPN11_somatic gene mutation compared to patients with other NF1, KRAS, NRAS, CBL, and PTPN11_germline gene mutations (median 34.0%, p = 2.29 × 10⁻⁶). 3 Furthermore, Figure 3A(h) shows that the proportion of abnormal monocytes of type D was significantly lower in the group of patients with the PTPN11_somatic gene mutation compared to patients with other gene mutations such as NF1, KRAS, NRAS, CBL, and PTPN11_germline (median 17.0%, p = 1.67 × 10⁻⁶). 3 ). Figure 3A(j) shows that patients with PTPN11_germline gene mutations had a significantly higher proportion of abnormal B-type monocytes compared to patients with other PTPN11_somatic, NF1, KRAS, NRAS, and CBL gene mutations (median 26.0%, p = 6.67 × 10⁻⁶). 3 ). Figure 3A(o) shows that the proportion of abnormal C-type monocytes was significantly higher in patients with KRAS gene mutations compared to patients with other PTPN11_somatic, NF1, NRAS, CBL, and PTPN11_germline gene mutations (median 41.0%, p = 3.65 × 10⁻⁶). 2 ). In the group of patients with NRAS gene mutations, the proportion of patients with type D abnormal monocytes was significantly higher compared to patients with other PTPN11_somatic, NF1, KRAS, CBL, and PTPN11_germline gene mutations (Figure 3B(t)) (median 48.5%, p = 1.03 × 10⁻⁶). 5 Figures 3B(q) and (s) show that the proportion of monocytes of types A and C is significantly lower (A: median 8.0%, p = 7.21 × 10⁻⁶). 5 C: median 13.5%, p = 9.67 × 10⁻ 4 ). From the results in Figures 3A and 3B, the distribution of monocyte morphology is, (1) If A is abundant and D is scarce, then it belongs to the PTPN11_somatic group; (2) If B is abundant, the PTPN11_GERMLINE group; (3) If C is abundant, the KRAS group; (4) If A and C are few and B and D are many, then the NRAS group; or (5) If the case does not fall into any of the above categories (1) to (4), the CBL group or NF1 group It was shown that it is possible to determine whether the patient has each of the following gene mutations.

[0066] Next, for types other than those for which the median values ​​were shown above, we calculated the median percentage of abnormal monocytes in all types A-D for each gene mutation group (PTPN11_somatic, PTPN11_germline, NF1, NRAS, KRAS, and CBL) for the 92 cases examined. The median values ​​for each type are shown in Table 2. Here, the sum of the medians for each group does not equal 100%, but we confirmed that when the cases in each group (for example, 27 cases of PTPN11_somatic, 2700 monocytes) are classified into A-D, the sum of each percentage equals 100%.

[0067] [Table 2]

[0068] The median values ​​in Table 2 were normalized according to formula (i) using the maximum and minimum values ​​for each type A, B, C, and D. Formula (i): (Median) - (Minimum) / (Maximum) - (Minimum) The normalized percentages of abnormal monocytes A through D were plotted with A on the positive y-axis, B on the negative y-axis, C on the negative x-axis, and D on the positive x-axis. Radar charts showing the distribution of monocyte morphology were then created for each gene mutation group of PTPN11_somatic, PTPN11_germline, KRAS, and NRAS. The results are shown in Figure 4.

[0069] From the results in Figure 4, the distribution of monocyte morphology is as follows: (1) If A and C are abundant and D is almost nonexistent, then it is the PTPN11_somatic group; (2) If A and B are abundant and D is almost absent, then the PTPN11_germline group; (3) If C is abundant and B and D are almost absent, then it is the KRAS group; (4) If A and C are almost absent and B and D are abundant, then the NRAS group; or (5) If a patient does not fall into any of the above categories (1) to (4), it has been shown that the patient may be determined to have a gene mutation in either the CBL group or the NF1 group, respectively.

[0070] For more details, see the entire radar chart. (1) The area occupied by the first quadrant is 0.0%; The second quadrant accounts for 88.5% of the area; The area occupied by the third quadrant accounts for 11.5%; and If the area occupied by the fourth quadrant is 0.0%, then it belongs to the PTPN11_somatic group; (2) The area occupied by the first quadrant accounts for 6.3%; The second quadrant accounts for 39.7% of the area; The area occupied by the third quadrant accounts for 46.7%; and If the area occupied by the fourth quadrant is 7.4%, then it belongs to the PTPN11_germline group; (3) The area occupied by the first quadrant accounts for 2.0%; The second quadrant accounts for 98.0% of the area; The proportion of the area occupied by the third quadrant is 0.0%; and If the proportion of the area occupied by the fourth quadrant is 0.0%, then it is the KRAS group; (4) The area occupied by the first quadrant is 0.0%; The area occupied by the second quadrant accounts for 0.0%; The proportion of the area occupied by the third quadrant is 0.0%; and If the area occupied by the fourth quadrant is 100.0%, then the NRAS group; or (5) If the case does not fall into any of the above categories (1) to (4), the CBL group or NF1 group It was found that this was the case.

[0071] When estimating genetic abnormalities in new cases using a blood smear, 100 abnormal monocytes are counted from the new case, and the proportions of A, B, C, and D are calculated to create a radar chart as shown in Figure 4. The type of genetic mutation the new case has can be predicted by which of the radar charts (i) to (l) shown in Figure 4 the new case's radar chart approximates. If it does not approximate any of the radar charts, it can be predicted to be either the CBL group or the NF1 group. Thus, (1), (2), (3), and (4) above are determined by referring to the radar chart shown in Figure 4. [Explanation of symbols]

[0072] 1 Judgment device 9 Network 10 Data Processing Unit 11 Acquisition Department 12 Extraction part 13 Classification section 131 Nuclear Morphology Classification Department 132 Cytoplasmic vacuole classification department 133 Type Classification Section 14 Distribution calculation part 15 Judgment section 16 Output section 20 Auxiliary storage 21 Blood smear image data 22 Abnormal Monocyte Image Extraction Parameters 23 Abnormal Monocyte Classification Parameters 29. Judgment Program 31 Input section 32 Display section 33 Communication I / F Section 33 Communication Interface Section 90 Server Equipment 100 Judgment System

Claims

1. A method for determining gene mutations in juvenile myelomonocytic leukemia (JMML), The distribution of monocyte morphology in blood samples taken from the subjects was: (1) If A is abundant and D is scarce, then the PTPN11_somatic group; (2) If B is abundant, the PTPN11_germline group; (3) If C is abundant, the KRAS group; (4) If A and C are few and B and D are many, then the NRAS group; or (5) If the patient does not fall into any of the above categories (1) to (4), the patient will be classified into the CBL group or the NF1 group. The process includes determining whether the subject possesses each of the following gene mutations: The aforementioned A is a monocyte having a cytoplasmic vacuole and a club-shaped nucleus; The aforementioned B is a monocyte that does not form a cytoplasmic vacuole and has a club-shaped nucleus; The aforementioned C is a monocyte having a cytoplasmic vacuole and a round nucleus; and The aforementioned D is a monocyte that does not form a cytoplasmic vacuole and has a round nucleus. A method to demonstrate each of these.

2. If (1) above has a large amount of (1')A and C and almost no D, then it is the PTPN11_somatic group; If (2) above contains a large amount of (2')A and B, and almost no D, then the PTPN11_germline group; If (3) above is a group where (3') C is abundant and B and D are almost absent, then it is the KRAS group; If (4) above is the case where (4')A and C are almost absent and B and D are abundant, then it is the NRAS group; or If (5) does not fall under any of (1') to (4') above, then the CBL group or NF1 group. The method according to claim 1, which is a step of determining that the subject has each of the gene mutations.

3. The method according to claim 1 or 2, wherein, in (1) above, if the radar chart obtained by plotting the normalized ratios of A in the positive y-axis direction, B in the negative y-axis direction, C in the negative x-axis direction, and D in the positive x-axis direction approximates a radar chart where A=1.00, B=0.13, C=0.82, and D=0.00, it is determined to be the PTPN11_somatic group.

4. The method according to claim 1 or 2, wherein, in (2) above, if the radar chart obtained by plotting the normalized ratios of A in the positive y-axis direction, B in the negative y-axis direction, C in the negative x-axis direction, and D in the positive x-axis direction approximates a radar chart where A=0.85, B=1.00, C=0.38, and D=0.06, it is determined to be the PTPN11_germline group.

5. The method according to claim 1 or 2, wherein, in (3) above, if the radar chart obtained by plotting the normalized ratios of A in the positive y-axis direction, B in the negative y-axis direction, C in the negative x-axis direction, and D in the positive x-axis direction approximates a radar chart where A=0.50, B=0.00, C=1.00, and D=0.02, it is determined to be a KRAS group.

6. The method according to claim 1 or 2, wherein, in (4) above, if the radar chart obtained by plotting the normalized ratios of A in the positive y-axis direction, B in the negative y-axis direction, C in the negative x-axis direction, and D in the positive x-axis direction approximates a radar chart where A=0.00, B=0.84, C=0.00, and D=1.00, it is determined to be an NRAS group.

7. The method according to claim 1, wherein a monocyte having a club-shaped nucleus means a monocyte whose maximum distance calculated by Convex hull is greater than the cutoff value calculated from the ROC curve used to create the model.

8. The method according to claim 1, wherein monocytes with cytoplasmic vacuole formation refer to monocytes in which the number of vacuoles identified by HoughCircles in OpenCV is greater than the cutoff value calculated from the ROC curve used to create the model.

9. The method according to claim 1, wherein the distribution of the monocyte morphology is obtained using at least 100 monocytes.

10. A device for determining gene mutations in juvenile myelomonocytic leukemia (JMML), The distribution of monocyte morphology in blood samples taken from the subjects was: (1) If A is abundant and D is scarce, then the PTPN11_somatic group; (2) If B is abundant, the PTPN11_germline group; (3) If C is abundant, the KRAS group; (4) If A and C are few and B and D are many, then the NRAS group; or (5) If the patient does not fall into any of the above categories (1) to (4), the patient will be classified into the CBL group or the NF1 group. The system includes a determination unit that determines whether the subject has the gene mutation: The aforementioned A is a monocyte having a cytoplasmic vacuole and a club-shaped nucleus; The aforementioned B is a monocyte that does not form a cytoplasmic vacuole and has a club-shaped nucleus; The aforementioned C is a monocyte having a cytoplasmic vacuole and a round nucleus; and The aforementioned D is a monocyte that does not form a cytoplasmic vacuole and has a round nucleus. A device that shows each of these.

11. A program for operating a computer as a determination unit of the apparatus according to claim 10.

Images