JPH03285696A - Method for testing susceptivility of carcinostatic agent - Google Patents

Abstract

PURPOSE:To surely measure changes in the multiplication of only cancer cells by selectively extracting the image signal of the cancer cells and colonies thereof from the image signals of a specimen obtained from an image and subsequently measuring the multiplication degree of the cancer cells from the image informations. CONSTITUTION:A collagen gel is employed as a substrate and an image analyzer is used as a means for measuring the multiplication degree of cancer cells. The image signals of the cancer cells and colonies thereof are selectively extracted from the image signals of a specimen obtained from an image. The multiplication degree of the cancer cells are measured from the obtained image informations of the cancer cells and the colony. In the measurement of the multiplication of the cancer cells, the number of the colonies is preferably counted from the image signals of the selectively extracted cancer cells and colonies. The volume of the colonies is preferably measured from the image signals of the selectively selected cancer cells and colonies.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to an anticancer drug susceptibility test method, and in detail,
The present invention relates to a method for testing the sensitivity of cancer cells to anticancer drugs by culturing cancer cells in vitro.

[Conventional technology]

To test the effectiveness or sensitivity of anticancer drugs, there are clinical test methods in which anticancer drugs are administered to cancer patients and diagnostic tests are conducted to detect changes in cancer tissue, animal experimental methods in which anticancer drugs are administered to animals transplanted with human cancer tissue, and After contacting cancer cells taken from a cancer patient with an anticancer drug, the cancer cells are cultured using an appropriate substrate, or the cancer cells are brought into contact with the anticancer drug during culture, and the proliferation rate of the cancer cells is measured after a certain period of time. There are cell culture test methods for measuring this.

At the initial stage of development of an anticancer drug, it is not possible to immediately apply clinical testing methods, so first, after fully confirming the anticancer effect of the anticancer drug on cancer cells, that is, the sensitivity to the anticancer drug, using cell culture test methods and animal experiment methods. , it is common to confirm the final efficacy using clinical testing methods. Animal testing methods have problems such as difficult management of experimental animals such as nude mice, a low transplant success rate for human cancer cells, and the need for a long period of time to confirm experimental results, whereas cell culture testing methods It has the advantage that it can be carried out using relatively simple equipment and operations, and results can be obtained quickly. Therefore, in the development of anticancer drugs, anticancer drug susceptibility testing using cell culture testing plays an extremely important role.

Traditionally, typical cell culture testing methods include:
HT CA (Human Tumor Clonog)
There is a method called the enic As5ay) method. In this method, a single cell dispersion is prepared from cancer tissue taken out from a living body, and after contacting this with an anticancer drug, the cancer cells are cultured in a soft agar medium, and after a certain period of time, the formed cancer cells are Sensitivity to the anticancer drug is tested by measuring the number of colonies and evaluating the rate of colony formation by the anticancer drug. Calculation of the number of colonies is performed visually or using a colony counter. Other methods include culturing using a monolayer culture method and examining the effects of anticancer drugs on cancer cells using isotope methods and DNA measurement methods.

[Problem to be solved by the invention]

However, the HTCA method, which is a typical anticancer drug susceptibility testing method in cell culture testing methods, has the following problems.

First, the culture method using soft agar medium has the advantage of suppressing the proliferation of fibroblasts, but because the colony formation rate of cancer cells is low, a large number of cells are required. fewer tests can be performed on collected cells. Moreover, there is a problem that the types of cancer cells that can be cultured are limited. That is, the types and properties of cancer cells vary depending on the tissue of the affected area, and among these many types of cancer cells, those that can be cultured on an agar medium are quite limited.

Therefore, the present inventors investigated a new anticancer drug susceptibility test in which soft agar medium was replaced with a collagen gel matrix in a primary culture system using human cancer cells. We found that even human cancer cells that do not have a
[Tissue Culture Research J Vol, 6. No, 1.
(see 1987). As a result, we were able to significantly expand the scope of application of cell culture testing methods. It was also found that collagen gel can culture cancer cells more efficiently than soft agar media, etc., so it has the advantage of requiring fewer cells and allowing test results to be obtained efficiently in a short period of time.

Note that when a collagen gel is used as a substrate, the degree of proliferation is generally measured by measuring the amount of DNA of cells proliferated within the collagen gel and calculating the degree of proliferation of cancer cells.

However, when cancer cells are cultured in a collagen gel matrix,
Depending on the type of cancer tissue, a problem arises in that fibroblasts contained in the cancer tissue proliferate at the same time as cancer cells.

Since fibroblasts contain DNA like cancer cells, the method described above for measuring cancer cell proliferation by quantifying DNA cannot be applied. Therefore, in this case, the formed colonies are measured visually or using a colony counter. However, when measuring visually using a stereomicroscope, it takes a lot of effort to pick out and measure only cancer cells, and there is variation among the measurers, making accurate measurement difficult. Also,
With a normal colony counter, it is virtually impossible to clearly distinguish and measure cancer cells and fibroblasts.

Among human cancer tissues obtained from clinical materials, many are cultured in a collagen gel matrix without separating cancer cells and fibroblasts because fibroblasts proliferate at the same time as cancer cells. It has been extremely difficult to test the sensitivity of cancer cells to anticancer drugs.

Conventionally, there are methods to suppress the proliferation of fibroblasts without suppressing the proliferation of cancer cells, such as the cisoxyproline method, the D-valine method, and the citrulline method. It can only be applied to fibroblasts and is difficult to put into practical use.

Therefore, this invention takes advantage of the collagen gel matrix, which allows cancer cells to be cultured well, and also prevents changes in the proliferation of cancer cells only, even if fibroblast proliferation occurs, which is the disadvantage of collagen gel matrix. An object of the present invention is to provide an anticancer drug susceptibility test method that allows simple and reliable measurement.

[Failure to solve the problem]

In order to solve the above problems, the present invention uses a collagen gel as a substrate to proliferate cancer cells, uses an image analysis device as a means to measure the degree of proliferation of cancer cells, and uses an image signal of a sample obtained by imaging. Provides an anticancer drug susceptibility test method for testing the sensitivity of anticancer drugs by selectively extracting image signals of cancer cells and their colonies from a computer and measuring the degree of proliferation of cancer cells from the obtained image information of cancer cells and colonies. do.

According to one aspect of the present invention, the degree of proliferation of cancer cells is measured by counting the number of colonies from image signals of selectively extracted cancer cells and their colonies.

According to another aspect of the present invention, the degree of proliferation of cancer cells is measured by measuring the volume of colonies from image signals of selectively extracted cancer cells and their colonies.

The collagen serving as the matrix may be any collagen that has gelling ability, but type II collagen with high gel strength is preferred. In addition, since optical measurements are performed,
Preferably, it has sufficient transmittance to light of the wavelength used for measurement in a gelled state and is optically homogeneous. Furthermore, it is desirable that the material does not undergo optical deterioration over time during culture.

Further, the step of culturing cancer cells on a substrate made of collagen gel is performed according to various conventional tissue culture methods using collagen gel. The timing of bringing anticancer drugs into contact with cancer cells is
For example, before seeding cancer cells (primary) from a patient on a collagen gel matrix, subculturing primary cancer cells and then seeding them on a collagen gel matrix, or
It may be set as appropriate, such as immediately after seeding on the collagen gel matrix or after culturing for several days (there is no particular limitation).

An image analysis device is used as a means of measuring changes in the proliferation of cancer cells. The purpose of an image analysis device is to clarify information about the object by connecting it to an optical device such as a microscope, converting the image signal obtained from it into numerical image information, and then performing arithmetic processing. used in To this end, the image analysis device is equipped with an image processing mechanism such as a microcomputer, an image storage mechanism such as a fixed disk device, and an output mechanism such as a monitor television or a video printer that outputs analyzed image information.

As the microscope, a microscope similar to a conventional microscope for visually measuring cancer cells can be used. Measurement ability or measurement accuracy can be improved by appropriately selecting conditions such as the magnification of the microscope, that is, the size of the field of view, the depth of focus, the object distance, the mounting structure of the culture container, the illumination device, etc. An input device such as a TV camera is attached to the microscope in order to input a sample image signal to the image analysis device.

Although it is possible to set the culture sample at the observation position of the microscope manually, it is possible to improve the efficiency of measurement by having a continuous transport mechanism that can continuously feed and take out the culture sample to and from the observation position. I can do it.

In this invention, a grayscale image of a culture sample is captured by a TV camera of an image analysis device. The captured sample images may include only images of cancer cells and their colonies (hereinafter collectively referred to as cancer cells), or may include a mixture of these images and the above-mentioned fibroblast images. The image analysis device removes unnecessary images of fibroblasts other than cancer cells from the sample image. To separate and remove images of cancer cells and other unnecessary images, use the shading and shape separation functions of the image analysis device. In other words, cancer cells are clusters of large numbers of cells that have a certain size, density, and outline, or form colonies, whereas fibroblasts, as their name suggests, have thin fibrous or They are branch-like, and cancer cells and fibroblasts can be clearly distinguished from each other by their shape and by differences in image shading. Therefore, image analysis equipment recognizes only images that meet certain conditions such as shape, area, density, and color difference as cancer cells, and recognizes images that do not meet the above conditions as unnecessary images such as fibroblasts. By doing so, images other than cancer cell images can be removed and only the remaining cancer cell images can be extracted.

By calculating and evaluating the shape of each image shown in this image of cancer cells, changes in the proliferation of cancer cells, that is, the degree of proliferation can be measured.

To evaluate the sensitivity of anticancer drugs, for example, cancer cells that have been separately contacted with one anticancer drug or two or more different anticancer drugs and control cancer cells that have not been contacted with the anticancer drug are cultured under the same conditions, and the What is necessary is to compare the changes in cancer cell proliferation measured by the image analysis device.

There are, for example, the following two methods for measuring the degree of proliferation of cancer cells from cancer cell images selectively extracted as described above. The first is to measure the number of colonies from images of cancer cells, and the second is to measure the volume of colonies from images of cancer cells.

According to the first method, for example, the number of colonies obtained by culturing cancer cells that have been brought into contact with an anticancer drug and the number of colonies obtained by culturing control cancer cells that have not been contacted with an anticancer drug are automatically measured and compared. That's what I do.

Further, according to the second method, for example, the volume of a colony obtained by culturing cancer cells that have been brought into contact with an anticancer drug and the volume of a colony obtained by culturing a control cancer cell that has not been contacted with an anticancer drug are calculated and compared. That's what I do. here,
Calculation of the colony volume is performed, for example, as described below, but is not limited thereto.

[For production]

By culturing cancer cells in a collagen gel matrix,
The number of types of cancer cells that can be cultured increases, the scope of application of cell culture testing methods expands, and since cancer cells grow well, test results can be obtained reliably and quickly.

1 2 By analyzing and measuring the amount of cancer cell colony formation using an image analysis device, it is possible to eliminate the influence of fibroblasts that proliferate at the same time as cancer cells. In other words, in images of cultured samples, cancer cells and fibroblasts have different shapes and concentrations, such as cancer cells being clumpy and dark, and fibroblasts being thin and fibrous and pale. There is a clear difference in terms of Image analysis devices can extract only images with shapes and densities that meet certain conditions from input images, so it is possible to separate cancer cell images and fibroblast images and extract only cancer cell images. I can do it. In addition to fibroblasts, images of unnecessary objects having shading or contours different from those of cancer cells can be separated from cancer cell images. By measuring the amount of cancer cell colony formation from the extracted cancer cell image, it is possible to measure proliferation changes of only cancer cells, not including fibroblasts.

For example, compared to the method of measuring with the naked eye under a microscope, measurements using an image analysis device can measure complex quantities much more efficiently and precisely in a much shorter time. Test results are stable because measurements can be performed under constant conditions, avoiding the possibility of variation in results or erroneous measurements.

Moreover, since it is a non-destructive measurement method, it is possible to measure the same sample over time, and when measuring the amount of change in the target object, that is, cancer cells, it is possible to easily achieve measurement accuracy that cannot be obtained with destructive measurement methods. . Furthermore, the steps necessary for analysis can be automated within the image analysis device, and by linking with a continuous measurement mechanism, the time and effort of measurement can be saved and streamlined.

When the degree of proliferation of cancer cells is measured by counting the number of colonies from selectively extracted images of cancer cells, the measurement can be performed with high accuracy using an inexpensive image analysis device if the colonies do not come into contact with each other.

In addition, when measuring the proliferation rate of cancer cells by measuring the volume of colonies from selectively extracted cancer cell images, it is possible to measure the number of colonies even if the colonies are in contact with each other or overlap. 3 More accurate measurements can be made compared to 4.

〔Example〕

Next, specific examples of the present invention will be described in detail, but the present invention is not limited to the following examples.

(Experiment 1) Comparison of Collagen Gel Culture with Other Culture Methods First, it was confirmed through experiments that the collagen gel matrix used in this invention is more suitable for culturing human cancer cells than the conventional soft agar medium.

Table 1 below shows the results of culturing various human cancer cells in a collagen gel matrix and in a conventional soft agar medium.

In the table, the culture method used in the conventional HTCA method was used for the soft agar medium. After culturing, observe the culture sample,
Cultivation was evaluated as successful if the number of cancer cell colonies exceeded 20. The numbers in each column indicate the number of successes/number of tests.

In addition, the number of seeded cells is 5 x 10' for the collagen gel matrix.
cells/ml, soft agar medium was 5×10′ cells/xl.

As can be seen from the above results in Table 1, it was demonstrated that far more types of cancer cells could be cultured by using the collagen gel matrix than by conventional soft agar media.

(Example 1) An anticancer drug susceptibility test of the present invention was conducted according to the following procedure, and the effects of multiple anticancer drugs were compared.

Culture of cancer cells First, cancer cells were cultured for image analysis. This was then embedded in collagen gel and cultured.

The embedding method is cell matrix type 1-A (0,3
% acid-soluble collagen solution: 8 volumes of Nitta Gelatin@), 1 volume of 10x concentrated Ham's F12 medium (NaHCO2 incompatible), reconstitution buffer (260mM-NaHCO
, and 50mM NaO containing 200mM HEPES
1 volume of FBS (fetal bovine serum) and 1 volume of FBS (fetal bovine serum) were added thereto, and the human cancer cells obtained by the enzyme treatment were added thereto, mixed well, and stored in water. This collagen mixture was dispensed into a Petri dish with a diameter of 35 u+ in 1 xi portions. Note that the number of cells was 5×10′ cells/i. By heating this to 37°C in a CO2 incubator, a collagen gel matrix containing human cancer cells was prepared.

Culture solution 1 containing an anticancer drug was added to the obtained collagen gel matrix.
layer and incubate at 37°C for 24 hours in a CO incubator.
The human cancer cells were allowed to come into contact with the anticancer drug by allowing the cells to stand for a period of time. Thereafter, the culture solution containing the anticancer drug was removed by suction, and then a culture solution containing no anticancer drug was added, followed by shaking in a CO2 incubator to wash the collagen gel matrix. This operation was repeated 10 times in total every 10 minutes to wash and remove the anticancer agent from the collagen gel matrix.

In addition, as a control, a culture solution containing no anticancer drug was used, and human cancer cells were not brought into contact with the anticancer drug.

Next, culture solution 2 was added and cultured for 10 days at 37°C in a CO2 incubator. In addition, the culture medium exchange is 1
I went every ~2 days. After culturing, the collagen gel matrix containing the cells was fixed using 10% formalin to prepare a sample.

The culture solution used above was DME + 10% FBS.
+In5ulin (1 μg/mf) +EGF medium (10 ng/mZ). Here, DME: Dulbecco's modified Eagle's medium (Dulbec
co's Modified [, agle's me
FBS: Fetal Bovine Serum (manufactured by Gibco) Insulin: Insulin (manufactured by Sigma) E
GF: Epidermal Growth Factor
.

(manufactured by Collaborative) were used respectively.

In addition, anticancer drugs include mitomycin (MMC: manufactured by Kyowa Hakko Kogyo), fluorouracil (5FU: manufactured by Kyowa Hakko Kogyo), and vindesine sulfate (VDS: Shionogi & Co., Ltd.).
), etoposide (vp16: manufactured by Nippon Kayaku@), and cisplatin (CDDP Nibris l ~ manufactured by LeMyers)
Five types were used.

The contact concentration of the anticancer drug to the cells was 1/10 (concentration in tumor tissue) of the highest clinically used concentration in the blood of each drug. That is, MMC = 0.1 μg / 11', 5
F U = 1.0, ij g /III! , V
D S = 0.00 5//g/+aL VP
-16=0.117g/+uZ, CDD P
= 0.2 μg/1/.

Image analysis processing The morphology of the colonies of human cancer cells proliferated within the collagen gel matrix of the fixed sample obtained as described above and the morphology of the fibroblasts proliferated at the same time was photographed with a TV camera through a stereomicroscope, and the obtained images were Image analysis device (LUZEXI)
IIU (manufactured by Nikon) as an image signal.

The specifications of the image analysis device are as follows.

Control processor: Intel 80386/732 bit processor Storage device: 40 MB Fixed disk device Image processing circuit: Image capacity maximum 1024 x 1024 pixels x (8
+1) Bit image array processor Image input device: TV camera, optical image pickup tube, optical microscope 90 Image output device: Digital/analog RGB monitor video printer The image signal input to the image analysis device is decomposed into pixels, The shading corresponding to each pixel of the image was digitized, and processing was performed to separate images of cancer cell colonies and fibroblasts from the original image. In the process of object image separation, first, the density features of the object were emphasized through logical calculations. The calculation used a combination of high frequency component extraction calculation, concentration class division calculation, and noise removal. From the obtained enhanced image, only the image of the object was extracted according to a certain standard value regarding density.

At this stage, the number of colonies containing only human cancer cells was counted based on the difference in density or the rate of change in density. In the image after image processing, only the target cancer cell colonies are emphasized, and unnecessary images such as fibroblasts are removed, so it is easy to count the number of colonies.

Verification of measurement results by image processing Compare the results of measuring colonies of only human cancer cells from the images after image processing as described above with the results of similar measurements performed without image processing. .

When a formalin-fixed sample that has not been exposed to anticancer drugs is viewed under a stereomicroscope, spherical cancer cell colonies and bipolar linear fibroblasts coexist.

In this state, we carefully measured the spherical colonies, that is, cancer cell colonies, with the naked eye, and found that they were 52/1.4 tm x 4° 4N (1 field of view).
Met. Note that the measurements were performed under high magnification using a stereomicroscope.

When the photographed image of the same sample was counted with a commonly used colony counter before image processing, the number of colonies was 1024/1.4 tm
1.4 m, and it was not possible to accurately count the number of cancer cell colonies due to the influence of fibroblasts.

Next, after performing the above-mentioned image processing (high-frequency component extraction calculation and concentration class division calculation), the number of colonies per 1 2 in the image is 51/1.4 x
1.4x*, and almost the same result as the measurement result with the naked eye was obtained. Therefore, it was demonstrated that the colony number measurement results obtained by image processing had sufficient accuracy. Additionally, it is much easier to measure the number of cancer cell colonies in an image after image processing than to measure only the number of cancer cell colonies from a mixture of fibroblasts, etc. using a microscope. Met.

Colony count measurement results after image analysis - Table 2 below shows the results of image analysis and colony count measurement as described above for samples contacted with each anticancer agent. However, the number of colonies was measured under low magnification (one field of view 2.8 m
2.8 cranes).

Table 2 In the above table, (-) indicates the case where no anticancer drug was used (control), and compared to this control, the anticancer drug of the sample with a small number of colonies, that is, the sample in which the proliferation of cancer cells was suppressed. Yes, it means that you are highly sensitive.

(Experiment 2) Comparison with the conventional method In order to confirm that the above-mentioned test method of the embodiment of the present invention shows good sensitivity to anticancer drugs, it was confirmed that it currently has the best clinical correlation in terms of antitumor effects. Comparisons were made with results measured using the nude mouse method.

First, from the colony number measurement results obtained in the above example, the percentage of the number of colonies when treated with an anticancer drug relative to the number of colonies when no anticancer drug treatment was performed (control) was calculated as the percentage of control (control). pe
% of control).

The results are shown graphically in FIG.

Next, an anticancer drug susceptibility test (Comparative Example 1) using nude mice was conducted as follows. Said "Culture of cancer cells"
The nude mouse transplantation system human colon cancer used in this study was cut into approximately 2 square pieces and transplanted subcutaneously into nude mice. Thereafter, when the estimated tumor weight reached approximately 100 cm, the MMC-3
/kg, 5-FU=50nv/kg, VP-16=2
5m1r/kg, CDDP=5■/kg or VDS
=3■/kg was intraperitoneally injected into each group of 5 nude mice for a total of 4 times every week. Also,
Another group was used as a control to which no anticancer drug was administered. The estimated weight of the tumor is determined by the major axis (L 111 ) and the minor axis (W
+i+s) was measured and calculated from LXW''/2.

Then, one week after the final administration, the estimated weight of the tumor was measured again, and the case where anticancer drug treatment was not performed (control)
The percentage of the average tumor weight when each anticancer drug treatment was performed with respect to the average tumor weight was calculated as a percentage op control in the same manner as above.

The results are shown graphically in FIG.

Comparing Figure 1, which shows the measurement results of an example of this invention that combines collagen gel and image processing, with Figure 2, which shows the measurement results by the nude mouse method, it is found that there is an extremely good correlation for any anticancer drug. It was demonstrated that the method of the present invention is highly practical as an in vitro alternative to the in vivo nude mouse method. Furthermore, whereas the nude mouse method requires a test period of over a month, the example demonstrated that a highly accurate test could be performed in a short period of about 10 days.

When the degree of proliferation of cancer cells is measured by counting the number of colonies, the degree of proliferation may deviate from the accurate degree depending on the manner in which the cells proliferate. For example, as shown in Fig. 3, when six single cancer cells 1 are cultured in collagen gel 5, if the rate of cell proliferation is uniform and small as shown by arrow A, the cells will be similar to each other. 6 colonies 2 of size 2 are formed, and the colonies do not overlap each other. In this case, when the number of colonies is automatically counted, a result of 6 is obtained. However, when the degree of proliferation is uneven (as shown by arrow B in Figure 3)
Or, if the degree of proliferation is large (indicated by arrow C in Figure 3), multiple neighboring colonies 2 overlap, so
Originally, multiple colonies were determined to be one by automatic measurement.

For example, in the case of growth indicated by arrow B, the number of colonies is 4 in automatic measurement, and in the case of proliferation indicated by arrow C, the number of colonies in automatic measurement is 2 (in both cases, the exact number of colonies is 6). ). In order to resolve such errors, there are two ways to do this: ■ Lower the number of cells to be seeded (cell density) (for example, L x 10' cells/N1 or less), or ■ End the measurement before the colonies overlap. , etc. are possible. However, in many primary cancer cells, the cell density is 5 × 10' cells/mZ.
Since colonies cannot be observed unless cultured at the above-mentioned high density, the effect of the anticancer drug cannot be accurately determined by the method (2) above. In addition, since the degree of proliferation of cancer cells differs depending on the individual cancer cells, the above method (①) measures before the difference from the control and the difference in the effectiveness of different anticancer drugs are measured. It becomes impossible to accurately grasp the effects.

As described above, if the degree of proliferation of cancer cells cannot be accurately measured, inconveniences arise in actual anticancer drug sensitivity tests. For example, as shown in Figure 4, when six single cancer cells 1 are cultured in collagen gel 5, if they are not treated with anticancer drugs as shown by the arrows, six large colonies 2 are formed, but if three of these 7 8 overlap or are in contact with each other, the number of colonies is determined to be 2 by automatic measurement. On the other hand, when treated with an anticancer drug as shown by arrow E, if six small colonies 2 are formed and are separated from each other, the number of colonies is determined to be six by automatic measurement. Generally, the effect (proliferation inhibition degree) of an anticancer drug is calculated by the ratio of , and if the anticancer drug is effective, the ratio must be smaller than 100%. However, in the example shown in Figure 4, the ratio is 300% ((
6/2) x 100%].

When the degree of proliferation of cancer cells is measured by the number of colonies, it is contact and/or overlap between colonies that gives rise to such unreasonable results.

Therefore, in the present invention, it is preferable to measure the proliferation rate of cancer cells by measuring the colony volume. In this case, the effect of the anticancer drug (degree of growth inhibition) is calculated by the ratio of: If the anticancer drug is effective, the ratio is smaller than 100%. As shown in Figure 5, when adjacent colonies (for example, two colonies) 2 and 2 are in contact with each other, the volumes of these colonies are (
is the same as the total volume of the two (2) separated colonies. In other words, contact has little or no effect on the total volume of the colonies. Furthermore, even in the case of overlapping, the volume of the overlapping portions is very small, so it does not affect the overall volume much. Therefore, an accurate measurement of cell proliferation is always possible by colony volume, but not always by automatic counting of colonies.

9 0 As mentioned above, when measuring the degree of proliferation of cancer cells by image analysis, even if it is not possible to measure the degree of cell proliferation by counting the number of colonies, it is possible to determine the degree of proliferation by measuring the estimated volume of the colony. degree can be measured accurately. This was confirmed in the next experiment by comparison with the measured amount of DNA. In this experiment, unlike the previous example, a human cancer cell line containing no fibroblasts was used. Therefore, measurement of the degree of proliferation of human cancer cell lines can be accurately verified by measuring the amount of DNA.

(Experiment 3) Culture of cancer cells - Collagen gel embedding culture was performed using a human lung cancer cell line in the same manner as the control in Example 1 without anticancer drug treatment.

Image analysis processing - The human cancer cell line cultured in collagen gel was photographed in the same manner as above and input as an image signal to an image analysis device. The input original image was subjected to image analysis processing in the same manner as described above, and only images of cancer cell colonies were extracted (however, in this case, since fibroblasts were not included, their removal processing was not necessary). Next, the number of colonies was counted from this image in the same manner as described above, and at the same time, the integrated value of the three-dimensional volume estimated from the colony image was measured. The volume was measured by the following procedure.

1. Define a binary image that is a projected image of the colony as the original image.

2. A screen that can express different gradations is defined as an accumulation screen.

3. Copy the original image to the storage screen with a density value of 1.

4. Remove pixels forming the contour of the projected image from the original image.

5. Define the reduced projected image after removing the pixels forming the outline as a new original image.

6. Copy the original image to the storage screen and add a density value of 1.

76 Until the projected image disappears, 4. ~6. Repeat the process. As a result, a mountain-shaped three-dimensional image (not necessarily a real image) whose bottom surface is a surface surrounded by the contour line 1 2 of the first projected image is obtained.

8. The height is defined as the product of the density value accumulated in each pixel of the accumulation screen and the length of the pixel.

9. Integrate the height of each pixel on the accumulation screen. As a result, the volume of the three-dimensional image of the mountain priest is obtained.

10. Assuming that the actual colony is symmetrical with respect to the projected image, the volume of the colony is calculated by doubling the integrated value. This assumption was made because it is thought that in three-dimensional culture, the same shape that spreads horizontally will also spread vertically.

11. Output as necessary.

The number of colonies, colony volume, and DNA amount measurement results obtained from the above image analysis are summarized in Table 3 and FIG. 6.

3 4 From the results shown in Table 3 and FIG. 6, as shown by the DNA values, the cells proliferated according to the number of days of culture. The number of colonies measured up to the 5th day of culture is consistent with the change in DNA amount (although it peaks on days 5 to 7 of culture and decreases thereafter. This is due to cell proliferation. , as the cell density increases and the cells overlap, it seems that the accurate number of colonies is not being counted.On the other hand, the colony volume is based on changes in DNA values from the initial stage of culture to long-term culture. It can be seen that the volumetric method is suitable for measuring the degree of cell proliferation, especially for measuring the degree of proliferation over a long period of time.

Moreover, there are also the following problems in measuring the number of colonies. That is, the degree of proliferation of cancer cells generally differs depending on the individual cells (see arrow B in FIG. 3).

Since the colonies 21, 22, and 23 are different in size, the number of cells contained therein is naturally different, but when counting the number of colonies, each of the colonies 21, 22, and 23 is determined to be one. On the other hand, for one colony volume, two colonies
1.22 and 23 are calculated as different volumes.

(Example 2) Cancer cells were grown in the same manner as in Example 1, except that in "Culture of cancer cells" in Example 1, nude mouse transplanted human lung cancer was used instead of nude mouse transplanted human colon cancer. Culture was performed. The obtained sample was subjected to image analysis processing in the same manner as in Example 1, and only images of colonies were extracted.

From the obtained images, the number of colonies was counted in the same manner as in Example 1, and the integrated value of the colony volume was measured in the same manner as in Experiment 3 above.

(Comparative Example 2) In addition, in order to verify the results of Example 2, an anticancer drug susceptibility test using the nude mouse method was also conducted in the same manner as in Comparative Example 1 above using the same cells used in Example 2. Ta.

Each result was calculated as a percentage of control in the same manner as above, and is shown graphically in FIG.

5 6 As shown in the results shown in FIG. 7, there is an extremely good correlation between the result N of the nude mouse method and the result V of the colony volumetric method for all anticancer drugs. On the other hand, result C obtained by the colony counting method shows an abnormal value. This is because the cells used here have a high proliferation rate, and especially when not treated with anticancer drugs (control), the proliferated cells overlap as explained above, so the measured value of colony number does not accurately reflect the proliferation rate. However, colony volume measurements indicate that even in such cases, the degree of proliferation is accurately measured.

In other words, it has been demonstrated that for actively proliferating cells or when the cell density is high during culture, measuring the volume of the colony after proliferation is an effective method for measuring the degree of proliferation.

〔Effect of the invention〕

According to the anticancer drug susceptibility testing method of the present invention described above, by using collagen gel as a substrate for culturing cancer cells, many cancer cells that could not be cultured using conventional soft agar media etc. can be cultured. It becomes possible to do so.

By measuring the degree of proliferation of cancer cells using image analysis, it is possible to eliminate the influence of fibroblasts, etc. that proliferate together with cancer cells when they are cultured in a collagen gel matrix.
Test results can be obtained accurately and quickly.

When performing image analysis to determine the degree of proliferation of cancer cells, measuring the volume of colonies allows the degree of proliferation to be measured more accurately and over a long period of time than by counting the number of colonies.

As a result, compared to conventional testing methods, the range of application to cancer cell types is much wider, and the sensitivity testing method is suitable for the development of anticancer drugs because it is easy to operate and can quickly and accurately obtain test results. can be provided.

[BRIEF DESCRIPTION OF THE DRAWINGS] FIG. 1 is a graph showing the sensitivity measurement results of each anticancer drug in Examples according to the present invention, FIG. 2 is a graph showing the sensitivity measurement results of each anticancer drug in Comparative Examples, and FIG. teeth,
Proliferation in cancer cell culture 7 8

Claims (1)

[Scope of Claims] 1. A method for testing the sensitivity of anticancer drugs by growing cancer cells in an environment where a matrix exists and measuring the degree of proliferation, using a collagen gel as a matrix and measuring the proliferation of cancer cells. An image analysis device is used as a means of measuring the degree of cancer cells, and image signals of cancer cells and their colonies are selectively extracted from the image signals of the sample obtained by imaging. An anticancer drug susceptibility test method for measuring the degree of proliferation of 2. The anticancer drug susceptibility testing method according to claim 1, wherein the degree of proliferation of cancer cells is measured by counting the number of colonies from image signals of selectively extracted cancer cells and their colonies. 3. The anticancer drug susceptibility test method according to claim 1, wherein the degree of proliferation of cancer cells is measured by measuring the volume of colonies from image signals of selectively extracted cancer cells and their colonies.