JP2002521682A - Cell in situ analysis method - Google Patents

Abstract

  (57) [Summary] A method for in situ analysis of a biological sample, wherein (a) the biological sample is selected from the group consisting of a first immunohistochemical stain, a first tissue stain, and a first DNA ploidy stain. N types of stains, wherein the second stain is selected from the group consisting of a second immunohistochemical stain, a second tissue stain, and a second DNA ploidy stain A staining step, wherein N is an integer greater than 3, and (i) when the first staining agent is the first immunohistochemical staining agent, the second staining agent is The second tissue stain or the second DNA ploidy stain, and (ii) when the first stain is the first tissue stain, the second stain is the second stain. An immunohistochemical stain or the second DNA ploidy stain, while (iii) the first stain If the agent is the first DNA ploidy stain, the second stain is the second immunohistochemical stain or the second tissue stain, and (b) spectral data collection Collecting spectral data from the biological sample using an apparatus, wherein the spectral data collecting device and the N types of stains are capable of collecting spectral components associated with each of the N types of stains. Selecting.

Description

DETAILED DESCRIPTION OF THE INVENTION

TECHNICAL FIELD The present invention relates to in situ analysis of cells, and more particularly, to a plurality of immunohistochemical stains and tissue stains using a spectral imaging method, which enables not only high spatial resolution but also spectral resolution. And / or simultaneous in situ analysis of DNA ploidy stains.

BACKGROUND OF THE INVENTION Over the past 10-15 years, when applied to cells, immunocytochemistry (IC
Immunohistochemistry (IHC), also known as C), has become an essential tool in diagnostic pathology and has revolutionized surgical pathology practices substantially. The terms "immunohistochemistry" and "immunocytochemistry" are used interchangeably herein. Panels of monoclonal antibodies can be used to differentially diagnose undifferentiated neoplasms (eg, to distinguish lymphoma, carcinoma, and sarcoma) to reveal markers specific to certain tumor types or to identify malignant lymphomas. Diagnosis can be made, phenotyping, and demonstrating the presence of viral antigens, oncoproteins, hormone receptors, and growth-associated nucleoproteins. Not only are such markers important for diagnosis, but certain tumor markers have become prognostic and have been shown most widely in breast cancer. Examination of these markers can be performed by the IHC on a wide variety of subjects, including cytological samples, paraffin-embedded tissues and frozen sections. Most clinical assays use a single antibody for each slide, but the use of different colored chromogens allows the observation / declaration of more than one marker at a time. However, it is difficult to qualitatively or quantitatively detect these chromogens without an appropriate apparatus for processing the spectral resolution, since the absorption coefficients spectrally overlap.

For most applications of IHC, a qualitative assessment of the staining pattern is sufficient because the staining pattern is the total staining pattern of the intended tissue and cells that gives diagnostic information. The human visual system is skilled in such pattern recognition tasks and can be very subjective, but very quick. However, the human eye or CCD camera is completely incompatible with high spectral resolution. However, there are certain markers for which image analysis can play an important role. These include detection of hormone receptors, tumor growth markers, oncoproteins, tumor suppressor gene products (eg, p53), chemotherapeutic resistance factors, tumor angiogenesis, lymphocyte subset analysis and the like.

[0004] The general purpose of quantitative IHC (QIHC) is to allow targeted measurements of immunostaining reactions. While these measurements may be useful for collecting data for experimental purposes, the greatest potential for such measurements is to assist in determining the diagnosis, prognosis, or treatment of human disease, especially cancer. To find out their use as Aside from the mechanics of making such measurements (discussed below), a number of issues can affect how measurements should be made.
It is related to the specificity of HC.

[0005] Many biomarkers detected by IHC do not express as simple "all-or-nan", and the cell population under study can be quite heterogeneous in terms of positivity and staining. is there. Also, since some antigens can be expressed on the plasma membrane, in the cytoplasm or in the cell nucleus, localization of the staining reaction can be important. Therefore, both the distribution of markers in the cell population and the non-cellular localization need to be considered.

[0006] Unlike immunostaining with directly fluoresceinized antibodies, where the degree of staining is linearly related to ligand concentration, IHC is theoretically nonlinear for two reasons: (I) the enzymatic reaction forming the precipitation reaction product is non-linear,
ii) A wide variety of amplification methods using cross-linking or complexing reagents are typically used, but their purpose is to increase staining sensitivity. As a result, the amount of antigen cannot be calculated directly from the intensity of immunohistochemical staining. In addition, automation of the staining procedure provides greater consistency than is usually obtained using manual methods, but the quality of immunostaining may vary substantially from day to day.

[0007] However, some studies have shown that under certain conditions, a linear relationship between ligand concentration and immunohistochemical staining may be obtained (mostly using model systems). [Shown) [van der Ploeg M and Duij
ndam WAL, Matrix models: Essential too
ls for microscopic histochemical res
ear (matrix model: an essential means for microscopic histochemical research), Hi
stochemistry 1986; 84: 283]. It was suggested that the biotin-avidin method of staining may be less suitable for this purpose than the peroxidase-anti-peroxidase method. This is due to the drawback that the former is highly sensitive to low concentrations of antigen and that the higher the antigen concentration, the more steric hindrance the poor dynamic range. Rahier et al. [Rahier J, Stevens M, de, examining insulin levels in rat islet cells]
Menten Y and Henquin JC, Determination
of antigen concentration in tissue s
actions by immunodensitometry (measurement of antigen concentration in tissue section by immunodensitometry), Lab Invest
1989; 61: 357-363], using image analysis, could show a good correlation between radioimmunoassay and QIHC. Rahi
The key to the success of this method is the use of lower concentrations of antibodies, and er et al. make an important consideration that higher concentrations of antibodies do not show a linear relationship with insulin content.
It also indicates that higher concentrations of antibody must be used when the purpose of the measurement is to detect all cells containing the intended antigen, independent of level. However, higher spectral sensitivities are required when lower levels of antibodies are used, especially if multiple antibodies are to be used simultaneously.

[0008] It is clear that attempts to quantify the intensity of immunostaining require the use of calibration standards, possibly cell lines or some internal standard that expresses known levels of antigen. These problems are similar in each application of QIHC.

[0009] The most commonly used method in QIHC is by measuring population statistics, such as the percentage of cells that stain positive, rather than the absolute biochemical concentration of the target molecule. This method is similar to the method used for clinical flow cytometry, where the percentage of a cell population that expresses a given surface marker (rather than the expression level of that marker) is most commonly measured.

[0010] For a number of contemplated markers, such methods are biologically significant. For example, if you are interested in the expression of cell cycle related markers,
The percentage of cells expressing the marker, rather than the absolute level of expression, is determined to provide information about the proliferating fraction.

[0011] For certain markers, such as hormone receptor measurements in breast cancer, biochemical quantification, expressed as concentration, was performed by standard methods. However, tumors can show significant heterogeneity with respect to the expression of these molecules, and measurements that reflect this heterogeneity can provide more biologically relevant information. Eventually, the two tumors show the same biochemical level of a given molecule, and in fact, have a very different proportion of tumor cells that express at any level anyway. This problem appears to be best demonstrated for estrogen receptor measurements in breast cancer.

The ability to measure immunohistochemical reactions in tissue sections or cytological samples depends greatly on the choice of staining and imaging methods. Immunostaining (signal) levels must be detectable above background (noise) and must be distinguishable from other cell or tissue characteristics. In IHC, this is typically done by carefully titrating the antibodies (primary and secondary) to maximize specific staining and minimize background staining. The tissue is visualized using a tissue stain of a different color than the chromogen (also known as counterstain), so that the staining reaction can be easily visualized. For example, an immunoperoxidase reaction that produces a brown reaction product is typically combined with methyl green or toluidine blue tissue staining (counterstaining) to obtain visual contrast. However, when a plurality of immunohistochemical stains of different labels are used in combination with the tissue stains and / or the DNA ploidy stains, the marker stain is used so that each spectral component can be identified independently of the other spectral components. Discriminate between themselves and further counterstain or DN
It would be advantageous to be able to distinguish between the different stains used in the A-ploidy test.
This cannot be achieved by any of the conventionally known methods.

Although some image analyzers use true color analysis to measure specific regions of the color spectrum, most imaging systems are based on gray-scale (black and white) light measurement, and the measurement light It relies on an optical filter to select a specific wavelength. Stain combinations that exhibit minimal spectral overlap are currently most useful for QIHC. However, devices with increased spectral resolution allow the use of highly overlapping dyes.

[0014] CAS 200 System (Cell, Lombard, Illinois)
(Analysis Systems) uses two video cameras, one coupled to a 500 nm bandpass filter and the other coupled to a 650 nm filter. Tissues are immunostained using the immunoperoxidase method with diaminobenzidine (DAB) as a chromogen (brown reaction product) and methyl green as a tissue (nuclear) counterstain. Images of all nuclei of cells in the field of view were captured by one camera using a 650 nm filter, while other cameras captured images of the brown reaction product at 500 nm (maximal transmission of DAB and absorption of methyl green. Wavelength). The optical density of the reaction product is the light transmittance (
Gray level) can be obtained by converting the optical density into an optical density using a calibrated look-up table, and the relative area of the staining reaction (expressed by the area ratio stained with methyl green) can be calculated. A simple threshold method is used to determine the gray level for capturing cell nuclei and setting a lower detection limit for DAB based on slides stained with control antibodies. Combinations other than methyl green-DAB may be used, limited only by the overlap of the reagent spectra and the usefulness of the suitable filter combination. However, this system has inherent limitations. That is, only spectral components in which two spectra do not overlap can be detected. For devices with higher spectral resolution,
Multiple spectral components with greater spectral overlap between components could be detected simultaneously.

[0015] Standards have not yet been established in the field of QIHC for measurement, computer algorithms, staining methods or performance characteristics (accuracy, reproducibility) [Wells WA,
Rainer RO, Memoli VA, Equipment, standard
Rdization, and applications of image processing (devices for image processing, standardization and applications), Am J
Clin Pathol 1993; 99: 48-56]. Such standardization, although progressing, has been difficult to achieve in another technically relevant area, flow cytometry [Mujrhead K. et al. A. , Establ
issue of quality control procedure
s in clinical flow cytometry (establishment of quality control method in clinical flow cytometry), Ann NY Acad Sci 1
993; 677: 1-20; and National Committee fo
r Clinical Laboratory Standards, Clin
ical applications of flow cytometry;
Quality assurance and immunophenotyping
ing of peripheral blood lymphocytes;
tentative guideline (clinical use of flow cytometry; quality assurance of peripheral blood lymphocytes and determination of immunophenotype; preliminary guideline), Vilano
va: NCCLS; 1992; NCCLS document H42-T (I
SBN1-56238-15.5-5)]. Standardization issues that are important for clinical laboratory testing, including the exact antibodies and chromogens to be used, the number of fields and tissue sections to be measured, the units used to report results, quality control materials, and instrument calibration, are: , QIH
C has not been resolved yet. As is evident from the present description, such standardization is greatly advanced by the present invention.

[0016] Tumor cell growth is dependent on lymphoma [Braylan RC, Diamond LW,
Powell ML and Harty-Golder B, Percentage
of cells in the S phase of the cell
cycle in human lymphoma determined
by flow cytometry: Correlation with l
abeling index and patient survival (percentage of cells in S phase of cell cycle in human lymphoma as measured by flow cytometry: correlation with labeling index and patient survival), Cy
tomory 1980; 1: 171-174; and Bauer KD, M
erkel DE, Winter JN, etc., Prognostic impli
sessions of ploidy and proliferative
activity in diffuse large cell lymph
omas (prognostic significance of ploidy and proliferative activity in diffuse large cell lymphoma), C
ancer Res 1986; 46: 3173-3178], breast cancer [Clar
kGM, Dressler LG, Owens MA, Ponds G, O
ldaker T and McGuire WL, Prediction of r
elapse or survival in patients with
node-negative breast cancer by DNA f
low cytometry (prediction of regression or survival of node negative breast cancer patients by DNA flow cytometry), N Engl J Med 1989
320: 627-633; Silvestrini R, Daidone M;
G and Gasparini G, Cell kinetics as a pr
ognostic marker in node-negative bre
ast cancer (cell kinetics as a prognostic marker in node-negative breast cancer), Cancer 1985; 56: 1982-1987; and Si
gardson H, Baldetorp B, Borg A, etc., Indic
attors of prognosis in node-negative
breast cancer (prognostic indicator in node-negative breast cancer), N
Engl J Med 1990; 322: 1045-1053] and colon cancer [Bauer KD, Lincoln ST, Vera-Roman JM, et al.
Prognostic implications of prolifera
live activity and DNA annuity in
colonic adenocarcinomas (meaning proliferative activity and prognosis of DNA aneuploidy in colon adenocarcinoma), Lab Invest 19
87; 57: 329-335], and has proven to have prognostic significance for a wide variety of commonly occurring malignancies. In one study, tumor cell growth has independent prognostic significance even if total DNA content analysis ("ploidy") is not [Visscher DW, Zarb
o RJ, Greenawad KA and Crissman JD, Prog
nostic significance of morphological
parameters and flow cytometric DNA
analysis in carcinoma of the breast (
Morphological parameters and flow cytometric DNA in breast carcinoma
Prognostic significance of analysis), Pathol Ann 1990; 25 (1): 17
1-210].

Although mitotic counts are generally considered to be defective and unreliable measures of proliferation, they do not require any special preparation.

The incorporation of radiolabeled thymidine, the “thymidine labeling index” (TLI), is a well-established method and has been described by the majority as the “gold standard” for measuring tumor cell kinetics. I have. Although TLI gives an accurate assessment of S-phase activity and has histological correlation, this method is cumbersome and cannot be easily adapted to clinical experiments.

[0019] Flow cytometry (FCM) is primarily used for DNA content analysis ("ploidy").
It has been widely used to measure cell cycle activity by quantifying the phase portion.
However, this method has a number of significant technical limitations. First, it may be difficult to obtain a single cell suspension from solid tumors, and fluctuating tumor cell numbers may be lost during preparation. Second, tumor cells are differentially diluted by benign normal and inflammatory cells, which can cause an underestimation of the S phase fraction, especially for DNA diploid tumors. Third, the DN consisting of a series of overlapping curves
The complexity of the A content analysis ("ploidy") makes it impossible to accurately use the curve fitting algorithm to measure the S-phase portion of the histogram. Multicenter studies have shown that the reproducibility of the flow cytometry S-phase fraction is poor and the practical clinical usefulness of the measurement is somewhat questionable. Another problem associated with measuring cell kinetics by flow cytometry is that only the S-phase fraction is typically determined, while a significant proportion of the tumor cell population has entered the cycle but still synthesizes DNA. It appears to be the G ₁ phase of the cell cycle of cells not. Possibly, the two tumors have the same S-phase fraction, but differ significantly in the fraction of total cells in the non-resting phase, thus exhibiting different growth kinetics and responding to cycle-dependent chemotherapeutic agents May be.

For all of these reasons, the in situ method of tumor cell cycle analysis
It provides more biologically meaningful information than can be obtained using dissociated tumor cells [Weinberg I. et al. S. , Relative ap
privacy of image analysis and fl
ow cytometry in clinical medicine (relative suitability of image analysis and flow cytometry in clinical medicine), Bau
er KD and Duque RE, Flow cytometry: Prin
chips and applications (flow cytometry: principles and applications), Baltimore: Williams and Willkins; 1
992: 359-372; Weinberg DS, Proliferatio
n indices in solid tumors (growth index in solid tumors), Adv Pathol Lab Med 1992; 5: 163-19.
1]. In addition to ensuring that the measurements obtained are specific for tumor cells, the in situ method allows for more extensive sampling of tumors and measurement of tumor cell heterogeneity.

[0021] Quantification of nuclear DNA is becoming increasingly practical in both research and clinical applications.

Measurement of DNA content by flow cytometry or image cytometry is based on the assumption that the amount of stain represents the amount of DNA and that the amount of stain is accurately measured by an instrument. These assumptions are based on the assumption that (i) the DNA labeling method (fluorescent dye, dye reaction or staining) is specific (all DNA is labeled and only DNA) and stoichiometric (staining intensity is , Changes in proportion to DNA content) and is stable (staining intensity does not change over time and repeated measurements); (ii) light emitted by a fluorescent dye or absorbed by a stain The instrument used to measure the applied light is accurate (resulting in near true stain loading) and reproducible even if not close to true stain loading (
Which gives very similar measurements when repeated for the same nucleus),
And linear (a result completely proportional to the amount of stain is obtained).

Unfortunately, both the staining method and the prior art instrument have a well-known limitation that the final measurement does not represent the absolute amount of DNA actually present in the nucleus.

While stoichiometry is readily ascertainable by routine controls, specificity cannot be determined because there is no alternative to cytometric measurement of DNA content at the individual cell level.

For example, (i) a very specific fluorescent dye is GC (eg, mitramycin,
Chromycin A3) or AT (e.g., Hoechst, DAPI) binds to DNA base pairs and thus only partially detects DNA, resulting in base pair sequence dependent measurements; (ii) Dye reactions such as Feulgen involve acid hydrolysis,
Some of the DNA fragments are removed at the same rate as released from decondensed chromatin. Thus, these reactions only partially detect DNA and provide measurements that are proportional to the balance between intrinsic and heterochromatin; (iii)
The fixation medium will either inhibit or facilitate further hydrolysis, depending on the method of interaction with the histone.

As far as image cytometry of Feulgen stained nuclei is concerned, Beer-Lambert's law does not fully apply. This is due to the following reasons: (i) the linearity with respect to the optical density (OD) of the stain is progressively lost as the OD increases beyond 1 unit (this is the case for highly stained heterochromatin). (Ii) the distribution of the dye is not spatially uniform, thus resulting in an increasing distribution error as the pixel size increases beyond the resolution of the optical element; (iii) the dye The absorption coefficient changes with the wavelength of the light, and the calculation with the densitometer applies only to monochromatic light. Since they do not emit enough intensity for visual observation (unless combined with very large arc lamps), normal image cytometry does not use a monochromator and the center of maximum absorption is ±
Use a filter that is 10 nm (typically different for each device).

In addition to taking into account the above, the prior art instruments used with the sample are optically compromised, so that reflections, refractions and diffractions by glass slides, mounting media, lenses and prisms are not possible. Cause. Light that does not follow the expected geometric path will result in glare, also referred to as the Schwarzchild-Villiger effect, which will distort the ratio of the light intensity incident on the nucleus to the light intensity leaving the object. Therefore, all pixel OD calculations are slightly incorrect. This error increases as the optical field size increases. Thus, the use of a high performance microscope is essential to reduce this systematic error in densitometric measurements, which is the most important factor affecting the variability of image cytometry measurements.

Video cameras are also susceptible to vibration, electromagnetic fields, image afterglow and saturation, all of which affect the distortion of the signal before digitization. At present, there is no doubt that charge coupled device (CCD) cameras provide the most reproducible results as far as concentration measurements are concerned.

The above limitations all contribute to a widely investigated negative error. Thus, it is clear that true cellular DNA content cannot be obtained by cytometric measurements. Therefore, with some commercially available image cytometry instruments,
The availability of measurements in picograms of DNA per nucleus is extremely surprising, generally confusing, and deliberately misleading clinicians using such devices. .

[0030] Provided that the staining methods and assays are correctly performed and controlled, the quantification of DNA-specific stains can be interpreted in terms of total proliferative activity and total cellular genetic abnormalities, It provides a clear guide on how to proceed further with the intended tumor characteristics for differential diagnosis and prognosis.

Various stains and dye reagents for the quantitative staining of DNA have been recommended in the literature (see Table 1), but ultimately only one of them is DNA cell photometric. Globally accepted for measurement. This is a reaction named after Feulgen and Rossenbeck and is often referred to simply as Feulgen reaction. Strictly speaking, the Feulgen reaction is not a staining but a dye reaction. Terms such as Feulgen staining and Feulgen reaction are currently the only staining methods that are stoichiometric for DNA. This means that DNA staining is quantitative and specific. All other methods described in Table 1 are quantitative, but not DNA specific. The success of the Feulgen reaction for quantitative DNA staining is undoubtedly due to this uniqueness.

The Feulgen reaction is a complex cytohistochemical method consisting of various preparation steps.
Primarily, the reaction begins with a process called acid hydrolysis: a slide on which cellular material is immobilized is immersed in hydrochloric acid (HCl) to split the purine bases, adenine and guanine, from the DNA molecule, resulting in a purine-free solution. An aldehyde group is generated on the DNA-containing molecule (hereinafter referred to as aprinic acid (ATA)). In a second step, the slide is immersed in a Schiff reagent containing a dye that is covalently attached to the aldehyde group. After removing excess dye, slides are dehydrated and mounted as usual. With correctly stained material, cell nuclei are stained magenta, and cytoplasm and background are not stained. The various steps of the Feulgen reaction are critical if they must be standardized for good reproducibility of the staining performance. Therefore, potential error sources must be considered when considering the relevant preparation steps.

[0033]

[Table 1]

As usual, HCl is used for acid hydrolysis, but in principle any acid can be suitably used. Acids have two effects on DNA molecules: (i) remove purine bases to create aldehyde groups in DNA molecules; (ii) depolymerize large APA molecules into smaller fragments. These fragments are partially removed from the cell nucleus and diffuse into the acid solution. Nuclear staining intensity increases due to the formation of aldehyde groups, whereas A
Loss of PA fragments results in loss of stainable material from the cell nucleus, thus reducing staining intensity. In this way, we used two opposite chemical reactions. Each reaction follows a specific and complex kinematics, and the resulting reaction curve is called the hydrolysis profile or hydrolysis curve.

The hydrolysis profile can be further divided into four phases: (i) increased staining intensity; (ii) peak phase with maximum staining intensity; (iii) plateau phase with constant staining intensity; (Iv) Decrease in staining intensity.

[0036] Phase (i) is characterized by the continuous generation of aldehyde groups. AP
Loss of the A fragment is initially minimal, and the amount of stainable material in the cell nucleus increases steadily until phase (ii) is reached. The peaks are so short that they may not be visible in the hydrolysis profile. Looking at the balance between continuous generation of aldehyde groups and loss of APA fragments during the plateau phase (iii), it can be seen that the staining intensity remains constant over a period of time. In phase (iv), the loss of APA fragments outweighs the generation of new aldehyde groups and reduces staining intensity. After prolonged hydrolysis, the number of generated aldehyde groups is maximum, but all APA fragments have been removed from the cell nucleus and the staining intensity is zero.

In practice, it is important to stop acid hydrolysis during the phase (iii), in which the smaller the fluctuation of the hydrolysis time, the smaller the influence on the staining intensity is.

The shape of the hydrolysis curve is affected by several factors, of which only some can be standardized.

Highly condensed DNA is less susceptible to acid hydrolysis than decondensed DNA: phase (i) and phase (iv) have less steepness and the plateau phase is delayed. The compactness of DNA as a biological property of a cell varies depending on the type of cell, and in the same cell varies during the cell cycle. Cell preparation methods, such as sample sampling, and especially the immobilization of substances, can artificially affect the compactness of DNA, and thus the acid sensitivity.

When the acid bath is hot, all four phases are shortened. Phase (iii) can be as short as a few seconds and a significant plateau may not be detectable. If the hydrolysis peaks at a very steep phase (i) or (iv), the variation in the staining will be considerable, even if the processing time varies even in a very short time. A short hydrolysis profile with a very short peak phase is usually seen in the so-called thermal hydrolysis method at 60 ° C. using an acid bath and it is not possible to stop the hydrolysis at the right time, ie phase (iii). Often it is possible. More often, 4 to
Cold hydrolysis with 5 mol / l HCl is carried out at about 22 ° C. in this case,
Under normal conditions, the length of the plateau phase is a few minutes. Next, the slide is rinsed with tap water to stop acid hydrolysis.

Although various additives to the acid bath have been recommended in the literature, they must minimize the loss of APA fragments from the cell nucleus. However, these additives do not play an important role in practical use.

[0042] Schiff reagents are colorless aqueous reagents containing a dye mixture called basic fuchsin. Basic fuchsin is usually composed of four cationic triarylmethine dyes (pararosaniline and its methylated homologs rosanilin, magenta II and new fuchsin). High quality basic fuchsin contains a high proportion of pararosaniline. Schiff's reagent is colorless because the related dye exists in a leuco form with sulfite bound to the dye molecule. The color of the Schiff reagent (based on basic fuchsin) indicates loss of sulfite and degradation of the solution, which must then be discarded. A variety of alternatives to basic fuchsin have been recommended, among which is the thiazine dye thionine. The advantage of thionin is that it stains the cell nucleus blue, a color that cytologists and pathologists use to visually check the slides, making it easier to counterstain the cytoplasm with eosin Y or Congo red. realizable. Thionine Schiff reagents are usually not completely colorless.

After staining with Schiff reagent, the material is rinsed with dye-free sulfite water. Sulfite removes excess dye from the nucleus and cytoplasm, leaving only the covalently bound dye molecules immobilized on APA molecules in the nucleus. The background of the slide must be completely unstained when the Feulgen reaction is performed correctly.

[0044] Acid hydrolysis is the most important step in the Feulgen reaction. Properly performed hydrolysis must stop in the plateau phase. It is important to use a suitable molarity of HCl at the correct temperature (eg, 5 mol / l H at 22 ° C.).
Cl or 4.0 mol / l HCl at 27.5 ° C). Suitable molar acids are
It may be a commercially available product or can be easily prepared from concentrated HCl. Often, HCl stored in a 4 ° C. refrigerator is used immediately without waiting for the acid to warm up, which hinders the hydrolysis reaction and stops the reaction before it reaches the plateau phase. There are often. It is recommended to use a temperature controlled water bath. If this is not possible, measure the temperature of the acid solution carefully to help avoid erroneous photometric measurements.

It is very important to realize that changing the fixative can significantly affect the hydrolysis profile (by altering the chromatin compactness and thus the acid sensitivity of the DNA). Changing the fixative means that the hydrolysis curve must be re-evaluated. Therefore, it may be better to fixate in one staining protocol that has been established in relevant experiments and found to yield reliable and consistent staining results.

The staining method itself is not important. Staining must be performed for at least 45 minutes to allow enough time for the reaction to be completed. High quality and consistent quality Schiff reagents are commercially available.

Galocyanine chrome alum (GCA) is a cationic oxazine dye that forms a complex with a metal. GCA quantitatively stains DNA and RNA,
Not NA specific. Photometric measurements of DNA require luminosity correction for stained RNA or enzymatic or hydrolytic removal of RNA. Einarson
The GCA staining protocol defines a staining time of up to 48 hours at elevated temperatures and cannot be used for routine cytology. Hustain and Watts
Proposed a modified GCA with a staining time of about 15 minutes.

Compared to Feulgen, GCA after Hustain and Watts has the following advantages: (i) no acid bath; (ii) short staining time of 15 minutes; (iii) gray- Blue cell nucleus. On the other hand, disadvantages are (i) not DNA specific; (ii) there is background staining (by RNA); (iii) the shelf life of the staining solution is short (about 6 weeks). The specificity of the staining is determined by mild hydrolysis (1 mol / l HCl, 22 ° C., which removes RNA but not DNA).
, 10 minutes). Anion counterstaining can be performed without loss of GCA from the stained DNA.

GCA stains both wet and dry fixatives. Dry fixed slides have significantly lower staining intensity. Preferred fixatives are 99% (v / v) ethanol for 10 minutes for wet fixatives or 3.7% (v / v) neutral buffered formaldehyde for dry fixatives. If using a commercial spray fixative containing polyethylene glycol (PEG), remove the PEG film on the slide and then stain the slide by washing it with 99% (v / v) ethanol for 5 minutes. There is a need to.

Both methods (Foilgen reaction and GCA) are equally suitable for the cytometric measurement of DNA. However, when nuclear chromatin tissue is measured with high optical resolution,
The results of these methods are completely different: acid hydrolysis significantly alters the chromatin structure, and the chromatin tissue of the Feulgen stained cell nuclei is completely different from the tissue of the GCA stain.

GCA is generally less critical than Feulgen's reaction and easier to perform. Nevertheless, the present inventors consider the Feulgen reaction to be preferable in terms of substrate specificity and stability of the staining solution. However, constant staining quality requires careful standardization of the protocol.

[0052] A wide variety of monoclonal antibodies to cell cycle-associated nuclear antigens are commercially available and can be used for immunohistochemical staining of tissues and cells. The number of publications from studies on human tumors using these antibodies is increasing exponentially.

The most commonly used antibody is Ki-67, which stains proliferation-associated nuclear antigen in all lineages of human cells [Gerdes J, Sch
wab U, Lemke H and Stein H, Production of
a mouse monoclonal antibody reactive
e with a human nuclear antigen assoc
iaed with cell proliferation (production of mouse monoclonal antibody reactive with human nuclear antigen associated with cell proliferation), Int
J Cancer 1983; 31: 13-20; Gerdes J, Lemk.
e H, Baisch H, Wacker H-H, Schwab U and St
ein H, Cell cycle analysis of a cell
provisionation-associated human nucleus
ar antigen defined by the monoclonal
anKi-67 (cell cycle analysis of cell proliferation-related human nuclear antigen as defined by monoclonal Ki-67), J Immunol 1984; 133: 17.
10-1715; and Gerdes J, Li L, Schlueter C
Etc., Immunobiochemical and molecular bi
organic characterization of the cell
provisionation-associated nuclear ant
ien that is is defined by monoclonal a
antibody Ki-67 (immunobiochemical and molecular biological characterization of cell proliferation-associated nuclear antigen as defined by monoclonal antibody Ki-67), Am J Pa
thol 1991; 138: 867-873]. This antibody has the useful property of staining the nuclei of cells in the G ₁ , S, G ₂ and M phases of the cell cycle, but not the nuclei of resting (G ₀ phase) cells. Several studies have shown that the percentage of tumor cell nuclei stained by the Ki-67 antibody was similar to breast carcinoma,
It has been shown to correlate with tumor grade and other prognostic features for a wide variety of tumor types, including lymphomas, meningiomas, glial and astrocytic brain tumors, malignant melanomas and sarcoma. In non-Hodgkin's lymphoma, Ki-67 staining is associated with the class of action, and the increase in the proliferating fraction as measured by Ki-67 staining is associated with an incorrect prognosis. In breast cancer, Ki-67 staining was correlated with nuclear grade and lymph node status, and several studies have shown the prognostic significance of Ki-67 staining in this type of cancer. The clinical utility of the Ki-67 antibody was somewhat limited by the antigen being stored in frozen tissue and lost in standard fixation. However, a monoclonal antibody (MIB1) was produced that appeared to stain the same antigen as that stained by Ki-67, and could be used on paraffin sections. Thus, extensive retrospective studies on preserved paraffin-embedded tissues could be performed to determine the prognostic significance of staining with this antibody. The recently described antibody Ki-S1 appears to have similar staining properties to Ki-67 and can be used in paraffin sections. The relative fraction of tumor cell nuclei positive for Ki-S1 staining appears to have prognostic significance in breast cancer.

[0054] Other proliferation-related nuclear markers can be shown to have clinical utility.
PCNA / cyclin is a 36 kDa nucleoprotein present in proliferating cells,
It is an auxiliary protein of DNA polymerase-Delta. Flow cytometry studies revealed that two fractions of PCNA / cyclin were present in the cell nucleus. Among these, the fraction insoluble in the nonionic detergent is more limited to the S phase of the cell cycle. Immunohistochemical staining of this protein was demonstrated on frozen and paraffin embedded tissue fixed in alcohol or formalin. Therefore, research on preserved tissues is possible. However, staining of tissue sections depends not only on tissue preparation and fixation methods, but also on the antibody clone used, and the degree of antigen detection is susceptible to fixation duration, with longer fixation times resulting in greater antigen loss. Seem. It has been shown that the detection of PCNA in paraffin-embedded tissues can be improved using the microwave method, which provided excellent correlation with the flow cytometry S-phase fraction. Nuclei stained for PCNA / cyclin showed different positivity and studies showed that more strongly stained nuclei corresponded to S-phase cells. This is, of course, important for accurate quantification of antigen expression, independent of the method used.

Antibodies to alpha DNA polymerase have been applied to tissue sections of normal and malignant tissues and have shown clinical utility. P105 is expressed by the first _G 0 / _{G 1} phase transition, a nuclear antigen expression is increased through the M phase. However, attempts to use this antibody in tissue sections have been successful in many ways,
This proved to be a more useful tool in the study of flow cytometry than that of tissue sections. Although there is a change in the level expression of many oncoproteins during the cell cycle, the use of these markers in analyzing cell cycle activity in solid tumors is questionable, as overexpression is not completely relevant. There is room for Additional markers are summarized below.

In most immunohistochemistry studies using antibodies to proliferation-associated nuclear antigens, staining is measured by putative or time-consuming manual counting. The additional time required to perform such a manual analysis renders it unusable for normal clinical use, given the time limitations of busy pathology practices.

Quantification of staining for nuclear antigens by image analysis provides a means to obtain speed, accuracy and reproducibility of such measurements. The CAS 200 system described above uses two video cameras to view microscope images separately through optical filters. In this case, all of the ethyl green stained nuclei are captured in the visual field (650 nm), or the brown reaction product of the immunoperoxidase reaction (500 nm) is captured in the same visual field. By superimposing the map of the green stained area and the brown stained area, the calculation can be rapidly performed from the nuclear area% occupied by the immunohistochemical reaction. The threshold for positive staining is obtained using slides stained with a control antibody. This measurement is not the same as "% positive nuclei", which requires counting individual nuclei and correlating them with the staining reaction. Such measurements can be made more easily with cytological samples than with tissue sections where it may be difficult or impossible to distinguish individual nuclei that are often contiguous or overlapping. Nuclear area measurements were made quickly and the results are expressed as a cumulative average of all microscopic fields measured. The intensity of the staining reaction is not taken into account in this measurement. However, by adjusting the threshold value of positive staining, for example, only the most strongly stained nucleus can be measured instead of all the positive nuclei. However, this system is very limited in terms of spectral resolution and therefore has limited application. It will be seen that using a system that allows a similar combination of spatial decomposition and higher spectral decomposition will revolutionize the field of QIHC.

While performing QIHC, the choice of the field of view is important. This is because tumors can be quite heterogeneous in proliferative activity and it is not clear whether average or maximal proliferative activity is of most clinical significance. In examining malignant lymphoma, for example, Ki-67 staining was measured using random field selection (computer generated) or a selective measurement of the field determined to have the largest growing fraction. Measurements by random visual field selection are not only available for lymphoma grade, but also for flow cytometry S
While it was found to correlate well with the phase fraction, those selecting only the region of greatest growth lost these correlations and did not differ between grades. Therefore, for malignant lymphomas, measuring the average proliferative activity over multiple fields of view tends to be the most informative. In solid tumors with varying amounts of stroma, it makes sense to exclude stromal and inflammatory cells from the measurement. Studies on the measurement of Ki-67 expression in breast cancer showed a relatively weak correlation with the flow cytometric S-phase fraction, indicating that various dilutions of tumor cells not occupied by flow cytometry were observed. Given that it is predictable.

In order to obtain a statistically reliable measure of the average Ki-67 staining, it is necessary to measure a sufficient field of view so that the change in the cumulative staining% does not differ by more than 5% of the average. is there. One investigator states that all such measurements are in% values +/- error limits (
Can be calculated based on the sample size). The number of fields depends on heterogeneity within the tumor, but statistical analysis as well as experience indicate that a field number of about 10-20 is sufficient to obtain a 95% confidence interval in most cases. Was done. However, this average measurement is not always the most biologically relevant.

[0060] Using image analysis for measurement of PCNA / cyclin staining in breast cancer, it was shown that there was a significant difference in the degree of tumor cell proliferation between the center and periphery of the tumor. Comparing the growth of in situ and invasive components in individual cases of breast cancer, these components usually show comparable Ki-67 staining, but have an effect on the classification of tumors with low or high growth fractions. It has been found that inconsistencies can occur. Assuming that osmotic components are more biologically relevant for the clinical process,
Careful attention to the histopathological aspects may be important in selecting the field of measurement. In addition, changes in the level and pattern of ongoing nuclear expression throughout the cell cycle were observed for Ki-67 and PCNA / cyclin. Using image analysis, not only tissue analysis, but also measuring lining within a limited intensity range, can provide measurements that are more closely related to specific phases of the cell cycle, such as the S phase. However, in each case, image analysis with good spatial resolution is very limited in terms of spectral resolution, and this limitation is solved by the present invention.

In many cases of breast cancer, responding to hormonal therapy, the presence of estrogen receptors associated with tumor cells has prognostic significance and may indicate predictive value in selecting patients for anti-estrogen therapy. Do you get it. Measuring tumor estrogen receptors most commonly by biochemical methods has become a standard in the diagnosis and treatment of breast cancer. Generally, 6 of patients with estrogen receptor positive tumors
0% to 70% respond to hormone therapy, whereas only 5% of receptor negative cases respond.

In the United States, biochemical assays for estrogen and progesterone receptors are used primarily for breast cancer, especially the dextran-coated charcoal (DCC) method. However, a number of technical problems associated with biochemical assays limit their usefulness. First, due to the widespread use of mammography, an increased proportion of resected tumors was too small for biochemical analysis (300-500 m
g of tissue is required), and the fine-needle aspirated sample cannot be evaluated by the usual method. At Brightham and Women's Hospital,
About half of resected breast tumors are too small for DCC analysis. Second, the lack of morphological correlation can cause significant errors in biochemical assays. This means that mixed benign breast structures can produce false positive results, and excessive tumor dilution by normal and inflammatory cells can produce false negative results.
Sampling errors can be difficult to detect unless frozen section preparation is performed on the tissue prior to processing (a step not commonly performed in most experiments).
Third, the DCC method is extremely labor intensive. Accordingly, there has been much interest in developing assays that avoid these problems and improve the prognosis and therapeutic utility of tumor hormone receptor measurements.

The availability of commercially available monoclonal antibodies to the estrogen receptor (ER) and progesterone receptor (PR) has allowed visual detection of hormone receptors using standard immunohistochemistry [Greene GL].
And Jensen EV, Monoclonal antibodies as
probes for estrogen redaction and
Characterization (monoclonal antibody as a probe for estrogen redetection and characterization), J Steroid Biochem 1
982; 16: 353-359]. Numerous studies have reported good correlation between immunohistochemical measurements of ER / PR and standard biochemical methods. Most of these studies use semiquantitative methods to analyze immunohistochemical staining in tissue sections, taking into account heterogeneous expression commonly found in tumors. For example, the HSCORE [MaCar] used by McCarty et al.
Use of am, such as ty KS, Szabo E, Flowers JL
onclonal anti-estrogen receptor ant
ibody in the immunohistochemical eva
lation of human tumors (use of monoclonal anti-estrogen receptor antibodies in immunohistochemical evaluation of human tumors), Canc
er Res 1986; 46: (Suppl.): 4244s-4248s].
Is obtained from the weight average of staining intensity and% of positive tumor nuclei, and requires a subjective evaluation of staining intensity. Using such a method, the sensitivity and specificity of the antibody method as compared to the biochemical assay was 88% and 94%, respectively (85). Other studies have yielded similar results.

Although most studies report optimal sensitivity using cryostat sections of frozen tissue, tests on cytocentrifuge samples, needle aspirates and paraffin sections have all yielded good results. . One study has shown that ER immunohistochemical assays are a more predictive clinical response to hormonal therapy than biochemical assays [Pertschuk LP, Kim DS and Nayer K.
Et al. , Immunohistochemical estrogen
and progesterone receptor assays in
breast cancer with monoclonal antib
odies: Histopathological, demographic, an
d biochemical correlations and relat
ionship to endocrine response and su
rvival (Immunohistochemical estrogen and progesterone receptor assay in breast cancer using monoclonal antibodies: histopathological, artificial and biochemical correlations and relationships to endocrine response and viability), Cancer 1990:
66: 1663-1670]. This is the strongest argument for a newer approach. More recently, some of the larger cancer research groups in the United States have allowed the use of immunohistochemical measurements of ER in clinical trials where this method would be more widely accepted.

A significant technical problem with the use of immunohistochemical ER results has been the difficulty in objectively analyzing and quantifying the staining reaction. Semi-quantitative methods such as HSCORE do not provide the reproducibility needed in experiments, and standards for measuring and reporting staining reactions have not yet been established. The use of image analysis offers the opportunity to obtain a standard of practice for the quantification of hormone receptors in breast cancer, and nuclear staining patterns can be easily adapted to QIHC as described above for the Ki-67 antigen, but with spectral resolution Is not so applicable to multiple markers / counter-stains.

The main intended characteristic to measure is the staining intensity as well as the percentage of tumor cell nuclei stained for ER / PR. Both% positive (using total nuclear area) and intensity were incorporated into quantitative immunohistochemistry (QIHC) scores, and excellent correlations between QIHC scores and ER biochemical measurements were reported. However, staining intensity is difficult to standardize and is subject to day-to-day variability in the experiment. The simpler approach used is to measure only the% staining nuclei area that appears to be linearly correlated with biochemical ER measurements of less than 200 fmol / mg protein. Using a 10% positive nuclear area cutoff, 92% sensitivity and 100% ER specificity compared to the DCC method.
%. Sklarew et al. [Sklarew RJ, Bodmer SC and Pertschuk LP, Comparison of microscopic
ic imaging strategies for evaluating
immunohistochemical (PAP) steroid r
acceptor heterogeneity (comparison of immunohistochemistry (PAP) microscopic imaging methods to assess steroid receptor heterogeneity), Cy
tomtry 1991; 12: 207-220] describes an imaging method for measuring ER heterogeneity in a tissue section in consideration of nuclear polymorphism and a change in a nuclear area caused by tissue section formation. Sklarew et al. State that this method provides an excellent predictor of response to endocrine therapy in breast cancer. Again, based on the imaging itself, this method is limited in terms of spectral resolution and is therefore not applicable to the detection of multiple immunohistochemical marker stains / tissue stains / DNA ploidy stains.

Since the treatment decision is based on whether the tumor is positive or negative and not on the absolute value, is it necessary to convert the hormone receptor measured by QIHC into a biochemical value? It is not clear yet. Indeed, since the presence of a significant tumor fraction that is negative for hormone receptors may predict that it will not respond to therapy, the tumor heterogeneity reported by QIHC may be at the biochemical level. It appears to have greater clinical relevance. Also,
Experience has shown that the majority of cases do not show tumor cells that are positive for ER (more than 50% cells positive) and few cases occur near the 10% positive cutoff. Thus, the continuum of values found in biochemical assays is
It appears to reflect more diverse dilution of tumor cells by benign tissue components than true biological range receptor expression. Consideration is needed to provide a standard method for reporting hormone receptor QIHC by comparing the clinical response to therapy with different image analysis features.

[0068] Tumor sampling is another important issue that needs to be addressed in breast cancer. More often than not, the frozen part of the tumor does not completely represent the whole tumor. For example, frozen tissue may contain only intraductal tumors, whereas invasive cancer foci can be found in permanent sections. The biological characteristics of invasive tumors are considered to be the most clinically relevant, and in situ and invasive tumor components are not expected to exhibit identical characteristics. The purpose of this study was to compare the invasive and non-invasive components of individual breast cancer cases with respect to hormone receptor expression using image analysis. Good agreement was found between these components for both absolute ER expression (positive, negative) and ER expression level (% nuclear area). However, there is a discrepancy in the PR representation, and thus sampling does not seem to be a significant issue for the ER, but does not seem to be the case for PR. Similarly, as reported by other investigators, differences were seen between in situ and invasive components of breast cancer with respect to oncogene expression. Therefore, it is necessary to determine the heterogeneity of expression for each marker and prepare guidelines for clinical measurement.

In theory, immunohistochemical reactions localized in the nucleus, cytoplasm or tissue can be quantified using image analysis. However, in the analysis of multiple markers, the spectral resolution must be much higher than would otherwise be possible with an imaging system at the same time.

The products of cell surface and cytoplasmic antigens, especially oncogenes, are of great clinical importance. For example, Slamon et al. [Slamon DJ, Godolphin W
, Jones LA, etc., Studies of the HER-2 / neu
proto-oncogene in human breast and o
Varian cancer (HER-2 / neu in human breast and ovarian cancers)
Proto oncogene studies), Science 1989; 235: 177-18.
2] have shown that both amplification and overexpression of the HER-2 / neu oncogene in breast and ovarian cancers correlate with incorrect prognosis. In this study, the intensity of immunostaining correlated well with other measures of gene expression, and expression levels in ovarian cancer correlated with survival. The authors of these studies state that measurement of oncogene products using immunohistochemistry has distinct advantages over biochemistry and appears to best correlate with clinical outcomes.

Various studies have shown the potential of QIHC for oncogene expression.
Czerniak et al. [Czerniak B, Herz F, Wersto R
P etc. Quantitation of oncogene products
by computer-assisted image analysis
and flow cytometry (quantification of oncogene products by computer-assisted image analysis and flow cytometry), J Histoch
em Cytochem 1990; 38: 463-466] uses image analysis to measure the cellular expression of some oncogene products expressed in the nucleus, cytoplasm or cell membrane, and obtains the results using flow cytometry. I found something similar to what I could get. Later, Czerniak et al. Used imaging to measure Ha-ras expression in bladder cancer [Czerniak.
niak B, Cohen GL, Etkind P, etc., Concurrent
Mutations of coding and regulatory
sequences of the Ha-ras gene in urin
ary blader carcinomas (H in bladder carcinoma)
a-ras gene coding and regulatory sequences), J Histoch
em Cytochem 1992; 23: 1199-1204], indicating a correlation between an increase in p21 expression and a specific gene mutation. Nuclear oncoproteins are
It can be measured using image analysis by a method similar to that described for the hormone receptor and the Ki-67 antigen, and can be expressed as a positive nuclear area% or a value obtained by combining staining intensity and positivity. This method is described in FIG.
J, Bakst G, Weisheit K, Solis O and Ross J
S, Image analysis quantification of imm
uno-reactive retinoblastoma protein
in human thiroid neoplasms with ast
reptavidin-biotin-peroxidase stainin
g technique (image analysis quantification of immunoreactive retinoblastoma protein in human thyroid neoplasms by streptavidin-biotin-peroxidase staining), Am J Pathol 1991; 139: 1213.
-1219] was used to measure retinoblastoma protein in thyroid tumors. However, again, in imaging analysis, the ability to do so makes QIHC more suitable for clinical use, but fails to detect multiple markers.

Due to the problem of controlling inter-day variability in staining intensity, QIHC may require the use of internal controls and calibration standards. Bacus etc. [Bacus SS, R
uby SG, Weinberg DS, Chin D, Oriz R and Ba
cus JW. HER-2 / neu one expression and
proliferation in breast cancers (HER-2 / neu expression and proliferation in breast cancer), Am J Pathol 199.
0; 137: 103-111] normalizes tumor protein measurements for DNA content for HER-2 / neu using a combination of Feulgen stain (to measure DN content) and immunohistochemical staining. It has become. Tumor cell lines using oncogenes of known expression levels were used for calibration, taking into account the daily variation in staining intensity. Therefore, the average concentration of tumor protein could be calculated per cell and expressed as a percentage level measured in a cell line known to show high levels of overexpression. The threshold for HER-2 / neu overexpression was determined based on the low expression levels present in normal tissues. Using this method, the authors were able to correlate oncogene expression in breast cancer with growth fraction growth factor expression and DNA content ("ploidy") [Bacus SS, Chin D, Stern RK, Ort
izR, Ruby SG and Weinberg DS, HER-2 / neu
one expression, DNA ploidy and prolife
eration index in breast cancers (HER-2 / neu expression in breast cancer, DNA ploidy and proliferation index), Anal Q
ant Histol 1992; 14: 433-445].

Another example of a cellular protein that may be important for treatment is multidrug resistance (
mdr) gene, P-glycoprotein, a product of cell surface-related ATPase that appears to be involved in one mechanism of non-specific chemoresistance. Grogan et al. [G
optimizat, such as rogan T, Dalton W, Rybski J
ion of immunohistochemical P-glycopr
otin assessment in multidrug-resist
ant plasma cell myeloma using three
antibodies (optimization of immunohistochemical P-glycoprotein evaluation in multidrug resistant plasma cell myeloma using three antibodies), Lab Invest 1
991; 63: 815-824] revealed a correlation between P-glycoprotein expression and regression in multiple myeloma and used image analysis to measure this protein detected by IHC. We have also found that QIHC is more sensitive than Western blots for detecting P-glycoprotein expression. Other markers of therapeutic resistance, such as glutathione transferase and topoisomerase II, could be considered as well.

Measuring tumor markers with prognostic and therapeutic significance has become an important part of diagnostic surgical pathology and cytology. Given the small size of many of the tumor samples (biopsies and needle aspirates), using in situ methods such as IHC, combined with analysis of measurements when quantification is needed, , It may be necessary to detect these tumor markers. For all laboratory tests, performance and performance standards must be developed for image analysis to ensure the accuracy and reproducibility of these measurements.

Similar to the immunohistochemical stain, the tissue stain was used for histopathology in a conventional manner, and was found to be useful for cancer evaluation.

The histopathology that underlies our understanding of disease is based on non-quantitative morphology using stained tissue samples where the formation of chromatin-stain complexes enhances the shape and structure of cellular and non-cellular components. It is evaluation. The nuclear structure appears in the isolated state as a dynamic reflection of the metabolic state of the nucleus and as a physical consequence of its total content of biochemical components. The major nuclear components are DNA, RNA, histones and non-histone proteins, minerals and water. These change during malignant transformation with cytoplasmic and functional dedifferentiation, increasing the nuclear / cytoplasmic ratio. Studies have shown that changes in nuclear structure, chromatin pattern and nucleoli size and number are morphological features of cancer diagnosis. Computerized karyotyping (CNM) of individual human breast cancer cells obtained by fine needle aspiration biopsy showed that the classification was more reproducible than subjective observations using normal histological samples. Quantitative estimates of various parameters, such as two-dimensional estimates of mean chromatin area in the nucleus, and the number of central chromatin regions have been shown to correlate with prognosis and discrimination in breast cancer. Mean nucleolus profile area and microscopic size of largest nucleoli were found to indicate prognostic value in ocular melanoma. In stellate gliomas, the relationship between optical density and nuclear profile area was found to correlate with patient survival.

The histopathological classification uses different staining protocols to highlight the cellular structure of tissues and cells. The most common staining methods used are hematoxylin and eosin (H & E) staining. Another method is to use P for fractionation of hydrocarbon components.
There are AS staining, Masson trichrome staining for extracellular staining of collagen and Romanovsky-Giemsa staining in hematopathology.

In a conventional H & E stain, counterstaining of cytoplasm and various extracellular substances is performed with eosin after hematoxylin staining of nuclei. In this process of nuclear staining, hematoxylin is oxidized to the purple dye hematein and has a net positive charge due to this metal salt. H & E staining methods have remained largely unchanged for half a century except that some of the steps have been automated. This means that the method is relatively quick, inexpensive, suitable for most situations, relatively easy to master, and most importantly, accurate for most subjects. This is probably due to the fact that a microscopic diagnosis can be made.

[0079] Additional methods have been used to solve certain problems, such as PAS staining and Masson's trichrome staining. In PAS staining, substances containing adjacent glycol groups or their amino or alkylamino derivatives are oxidized by periodic acid to dialdehydes. This dialdehyde combines with the Schiff reagent to form an insoluble magenta compound. Masson trichrome staining uses phosphotungstic acid or phosphomolybdic acid in combination with several anionic dyes such as basic fuchsin, light green and hematoxylin.

Romanovsky-Giemsa staining is routinely used in hematology practices and reveals a variety of hematopoietic cells, both normal and diseased. In the process of Romanovsky-Giemsa staining, acidic eosin binds hemoglobin (in red cells) and eosinophil granules (in granule cells), while basic azure binds chromatin and ribosome-rich cytoplasm. Following this phase, eosin and Azure B combine to form a complex, resulting in a purple component known as the Romanovsky-Giemsa effect. Structures exhibiting this effect exhibit a characteristic absorption band at about 550 nm, the so-called Romanovsky band. The Romanovsky-Giemsa method has also been used to identify various lymphoid elements (including mast cells) and microorganisms in paraffin-embedded materials.

In the assessment of invasive breast carcinoma, ductal and lobular carcinomas appear to have a similar histological appearance [Azzopardi JG, Ch
epic OF, Hartmann WH, Jafarey NA, Lomb
art-Bosch A and Ozello L (1982), The World
d Health Organization histologic t
yping of breast tumors (Histological typing of breast tumors by the World Health Organization), 2nd edition, Am J Clin Pathol 78: 8.
06-816]. Certain quantitative histopathological variables have been identified by morphological methods to aid in distinguishing between ductal and lobular carcinoma [Ladek.
arr M and Sorensen FB (1993), Quantitative
e histopathological variables in in
situ and invasive ductal carcinoma o
f the breast (quantitative histopathological variables in breast in situ and invasive ductal carcinoma), AMPIS 101 (12): 895-903.
]. Attempts to rank and identify tumors, or in other words, classify tumors, are based primarily on nuclear morphology and chromatin structure [Ladekarl M and So
rensen FB (1993), Quantitative histopa
thological variables in in situ and
invasive ductal carcinoma of the bre
ast (quantitative histopathological variables in breast in situ and invasive ductal carcinoma), AMPIS 101 (12): 895-903; Cornelis
se CJ, de Konig HR and Moolenaar AJ (1984)
), Image and flow cytometric analysis
of DNA content in breast cancer; rel
ation to estrogen receptor content a
nd lymph node involution (image analysis and flow cytometry analysis of DNA content in breast cancer; relationship between estrogen receptor content and involvement of lymph nodes), Anal Quant Cytol Hist
ol 4: 9-18; Stenkvist B, Westman-Naeser.
S and Holmquist J (1978), Computerized n
uclear morphology as an objective me
the for for characterizing human cancer
cell populations (computer-assisted nuclear morphology as an objective method for characterizing human cancer cell populations), Cancer Res
38: 4688-4977; Dawson AE, Austin RE and W
einberg DS (1991), Nuclear grading of
breast carcinoma by image analysis (nuclear grading of breast carcinoma by image analysis), Classificatio
n by multivariate and neural network
analysis (classification by multivariate and neural network analysis), Am J
Clin Pathol 95: S29-S37]. Morphological classification of other tumor types, including but not limited to leukemia, lymphoma, sarcoma and other carcinomas [eg, Clarke AM, Reid WA and Jack AS (19
93), Combined proliferating cell nucl
ear antigen and morphometric analysis
s in the diagnosis of cancerous lymp
H. infiltrates (combination of proliferating cell nuclear antigen and body type analysis in the diagnosis of cancerous lymphoid infiltrates), J. Am. Clin. Pathol. 46: 1
29-134] is also very widely practiced in both research and medical practice.

Nevertheless, as recently published after the NIH study group who assessed the reliability of histopathological diagnosis by the best pathologists in the field of cancer diagnosis, specialized pathologists are used to diagnose neoplasms. Are not even between. Based on this work group, histopathological decision-making is 100% subjective, independent of the origin of the subject, and the condition of this condition in histopathological diagnosis is limited to specific tumors It can be concluded that the method can be applied to differential diagnosis in all organs. These conclusions are based on Hu
In the editorial of man pathology 27: 1115-1116, A
"Discordan" by Bernard Ackerman (1996).
ce ammonium expert pathologists in diag
nosis of melanocytic neoplasm (disagreement among expert pathologists in the diagnosis of melanocyte neoplasms).

Almost 80% of breast carcinomas are glandular [Aaltomaa S and Lipponen P: Prognostic factors in bre
ast cancelers (reviews) (prognostic factors in breast cancer (report)),
Int J Oncol 1: 153, 1992; Toikkanen S and Jensuu H (1990), Prognostic factors an.
d long-term survival in breast cancer
rina a defined urban population (prognostic factors and long-term survival rate of breast cancer in a compartmentalized urban population), APMIS 98: 1005-
1014]. Discrimination between gland ducts and lobular carcinomas has been found to be useful in assessing patient prognosis and treatment decisions [Ellis IO, Galea M, B
Roughton N, Locker A, Blaney RW and Elsto
n CW (1992), Pathological technical f
actors in breast cancer: II Histologi
cal type; relation with with survival
in a large study with long term fall
ow-up (pathological prognostic factor in breast cancer: II histology; relationship to survival in large-scale studies with long-term follow-up), Histopathology 20: 4.
79-489; Eskelinen M, Lipponen P, Papina
ho S, Aaltoma S, Kosma VM and Klemi P (19
92), DNA flow cytometry, nuclear morp
hometry, mitotic indices and steroids
receptors as independent prognostic
factors in female breast cancer (DNA flow cytometry, karyotyping, mitotic index and steroid receptors as independent prognostic factors in female breast cancer), Int J Cancer
51: 555-561; and Toikkanen S and Jensuu H.
(1990), Prognostic factors and long-t.
erm survival in breast cancer in ad
defined urban population (prognostic factors and long-term survival of breast cancer in a compartmentalized urban population), APMIS 98: 1005-1014]. Tumors have certain differences in clinical behavior and metastatic patterns; lobular carcinomas are more focal and bilateral than ductal carcinomas [Azzopa
rdi JG, Chepick OF, Hartmann WH, Jafare
y NA, Lombart-Bosch A and Ozello L (1982)
, The World Health Organization histo
logical typing of breast tumours (Histological typing of breast tumors by the World Health Organization), 2nd edition, Am J Clin Pa
thol 78: 806-816], patient expectation of survival is better than normal [DiC
onstanzo D, Rosen PP, Green I, Frankli
n S and Lesser M (1990), Prognosis in inf.
Illustrating local carcinoma; an analy
sis of “classical” and variant tumor
s (prognosis of infiltrated lobular carcinoma; analysis of "classical" and mutant tumors), Am J
Surg Pathol 14: 12-23; du Toit RS, Loc
ker AP, Ellis IO, Elston CW, Nicholson
RI and Robertson JFR (1991), Anevaluatio
no of differences in prognosis, recur
rence patterns and receptor status b
etween invasive local and other in
volatile carcinomas of the breast (evaluation of differences in prognosis, recurrence pattern and receptor status between invasive lobular carcinoma of the chest and other invasive carcinomas), Eur J Surg Oncol 17: 2
51-257]. The two tumor types differ in morphology and the cells of the infiltrating lobular carcinoma are usually smaller than the ductal carcinoma, have less polymorphism, and have fewer mitotic figures. Infiltrating duct carcinoma cells have more raised nuclei [Azzopard
i JG, Chepick OF, Hartmann WH, Jafarey
NA, Lombart-Bosch A and Ozello L (1982), T
he World Health Organization histolo
medical typing of breast tumours (Histological typing of breast tumors by the World Health Organization), 2nd edition, Am J Clin Path
ol 78: 806-816].

[0084] Certain histological types of intraductal carcinoma were observed: comedo, phloem, micropapillary and solid. All of these are recognized and categorized by specific criteria, further subdivided mainly by architectural patterns, cell polymorphism and hyperpigmentation of the nucleus [Pa
ge DL and Anderson TG (1987). Diagnostic
Histopathology of the breast (diagnostic pathology of the chest). Edinburgh, Scotland: Churchill Livingsto
ne, 120-157; Lagios MD (1990), Duct carc
inoma in situ pathology and treatmen
t (in situ pathology and treatment of ductal carcinoma), Surg Clin
North Am 70: 853-871; and Lenington WJ
, Jensen RA, Dalton LW and Page DL: Ductal
carcinoma in situ of the breast: Het
erogenicity of individual lengths (in situ duct carcinoma of the chest: heterogeneity of individual lesions), Cancer 73
: 118-124, 1994]. The survival expectancy of lobular carcinoma is usually better than that of ductal carcinoma [DiConstanzo S, Rosen
PP, Green I, Franklin S and Lesser M (19
90), Prognosis in filtering lobula
r carcinoma: an analysis of “classica
l "and variant tumors (prognosis of invasive lobular carcinoma: analysis of" classical "and mutant tumors), Am J Surg Pathol 14
: 12-23; du Toit RS, Locker AP, Ellis IO
, Elston CW, Nicholson RI and Robertson J
FR (1991), An evaluation of difference
s in prognosis, recurrence patterns
and receptor status between invasive
lobbular and other invasive carcinom
as of the breast (evaluation of differences in prognosis, recurrence pattern and receptor status between invasive lobular carcinoma of the chest and other invasive carcinomas),
Eur J Surg Oncol 17: 251-257]. Lobular carcinomas are often more bilateral and multifocal [Ladekarl M and S
orensen FB: Prognostic, quantitative h
istopathological variables in logical
carcinoma of the breast (prognosis in lobular carcinoma of the chest, quantitative histopathological variables), Cancer 72: 2602, 19
93], the pattern of metastasis from the tumor was found to be different. Unfortunately,
The histological classification of breast carcinoma is poorly reproducible, and attempts to classify the morphological subtypes of lobular carcinoma with different prognosis would not be helpful [
Ladekarl M and Sorensen FB: Prognostic, q
positive histopathological variable
s in local carcinoma of the breast
(Prognosis in lobular carcinoma of the chest, quantitative histopathological variables), Canc
er 72: 2602, 1993]. Both the lobular and glandular types are now thought to arise from terminal gland-lobular units.

Characterization of nuclear features in different ways has been used for diagnosis, treatment and prognosis. Quantitative estimation of various histopathological parameters, such as two-dimensional estimates of nuclear profile area, nuclear profile concentration and mitotic profile number, was found to correlate with discrimination and prognosis. Changes in nuclear structure are a morphological feature of cancer diagnosis. Nucleus size, shape, and chromatin pattern have all been reported to be altered in breast cancer [Pienta KJ and Coffey DS: Correlat.
ion of nuclear morphology with prog
response of breast cancer (correlation between nuclear morphology and progression of breast cancer), Nuclear Morphometry of breast
canceler 2012, 1991]. However, heterogeneity of morphology and biology of tumors belonging to the same taxon was found to be the most prominent feature of breast cancer [Komitowski DD and Janson CP (1990), Qua.
nitative features of chromatin tru
cure in the promotion of breast can
cer (quantitative feature of chromatin structure in breast cancer prognosis), Cancer 6
5: 2725-2730].

For cytological specimens, staining patterns, nucleus size and total cell size, de-las-Morenas et al. [De-las-Mo
renas A, Crespo P, Moroz K and Donnelly MM
(1995), Cytologic diagnosis of duct
al vssus local carcinoma of the b
reest (cytological diagnosis of ductal carcinoma of the breast vs. lobular carcinoma), A
cta Cytol 39 (5): 865-869], eleven cytological parameters were found to be statistically different between infiltrating lobular carcinoma nuclei and invasive ductal carcinoma nuclei. Thus, the presence of coarse granular chromatin (nucleus size> 44 μm ² , cell size> 82 μm ² ) was found to be related to ductal carcinoma.

Ladekarl and Sorensen state that the major three-dimensional nuclear size, major nuclear profile area and mitosis index are all significantly greater in gland ducts than in lobular carcinoma, whereas the major nuclear concentration index is It was found small in ductal carcinomas [Ladekarl M and Sorensen FB:
Prognostic, quantitative histopatholo
gic variables in logical carcinoma o
f the breast (prognosis in lobular carcinoma of the breast, quantitative histopathological variables), Cancer 72: 2602, 1993]. Yu et al. Also identified several distinct nuclear features useful for distinguishing infiltrating duct carcinoma and lobular carcinoma [Yu GH, Sneige N, Kidd LD, Johnst.
on and Katz RL (1995), Image analysis der
inverted morphometric differences in fin
e needled aspirated of rectangular and lob
ullar breast carcinoma (difference in morphometry from image analysis of fine needle aspirated ductal carcinoma and lobular carcinoma), Anal
Quant Cytol Histol 17 (2): 88-92].

In addition to being used for solid tumors, histological stains and immunohistochemical markers can also be applied to the diagnosis, prognosis and treatment of various hematological malignancies, such as chronic lymphocytic leukemia ( CLL). Excessive accumulation of small mature appearance B lymphocytes in peripheral blood is a fundamental feature of the diagnosis of chronic lymphocytic leukemia (CLL). The appearance of mature lymphocytes with a sustained cell count of greater than 5 × 10 ³ cells / liter may be indicative of a transformation from normal to leukemia. The majority of normal lymphocytes and the predominant cell population of CLL are both composed of small cells with dense, aggregated nuclear chromatin that make it difficult to distinguish between these lymphoid populations by conventional light microscopy. ing. In some cases, the morphological characteristics differ from typical mature, small cell B-CLL: (i) a mixture of small and prolymphocytes (> 10% and designated CLL / PL) <55%), or (ii) CLL mixed with large lymphocytes.

The French-American-British group (FAB)
Criteria based on cytochemical and immunological methods have been proposed to establish a clear diagnosis. Immunophenotypic analysis indicates that large amounts of surface immunoglobulin (sIg) are present in normal B cells, which are weakly expressed or undetectable on B-CLL cells. CLL cells express the pan-B antigens CD19 and CD20 and the activation antigens CD5 and CD23, but do not express the terminal B cell differentiation antigens exhibited by plasma cells. B-CLL cells express kappa or lambda light chains, and monoclonality is essential to confirm the diagnosis. Mouse erythrocyte rosette (MRB
The receptor for C—R) can be detected on both B-CLL and normal B cells, but the two populations express different patterns of complement receptors, which can help distinguish them.

US Pat. No. 5,086,476 (Bacus et al.), Which is incorporated herein by reference, describes cells with a chromogen for proliferating substances and a counterstain for cell nuclei. An image processing method and apparatus for measuring the proliferation index of a cell sample by staining is taught. The chromogen is activated by the antibody-enzyme complex, which binds to the growth material and produces a stained cell sample. The stained cell sample is examined under a light microscope that forms part of the device to generate an enlarged cell sample image. The device optically filters the cell sample image and produces a pair of optically enhanced proliferator image and cell nucleus image. The enhanced image is analyzed electronically to determine the amount of cell nucleus and the amount of proliferating substance appearing in the image, respectively. The quantities are then compared to obtain a growth index for the portion of the cell sample that appears in the cell sample image.

US Pat. No. 5,109,429 (Bacus et al.), Which is hereby fully incorporated by reference, teaches a kit for quantifying components in cell nuclei. . The kit includes a stain and a microscope slide. Each slide is
It has a reference cell subject and a test cell area for accommodating test cells that are stained simultaneously with the reference cell subject.

US Pat. No. 5,202,931 (Bacus et al.), Which is hereby incorporated by reference in its entirety, describes an image analyzer for quantifying nuclear proteins in a cell population. Teaching. In particular, the hormonal receptor content of fine needle aspirates of human breast carcinomas has been evaluated. The estrogen receptor or progesterone receptor has been amplified and visualized in the test article by immunoperoxidase-type staining. A monoclonal antibody specific for the receptor,
After binding to the receptor site, it is amplified with a monoclonal antibody and a cross-linked antibody that binds to the peroxidase-anti-peroxidase complex. Chromogen (
Diaminobenzidine) is mixed with the complex, treated with hydrogen peroxide and reacted with the peroxidase to form an insoluble brown precipitate which marks the receptor site for optical identification. The test article is then counterstained with another chromogen (methyl green) that is specific for the nucleus of each cell. By two types of monochromatic light filtration, the region stained by the receptor site optical enhancer and the region stained by the nuclear region optical enhancer are optically separated. By measuring the optical density value of the stained receptor area, an intensity value is obtained that is directly related to the amount of hormonal receptor in the test article. Comparing the area of the nucleus containing the hormonal receptor to the total nucleus area gives a value (%) indicative of the distribution of cells throughout the receptor-containing test article. The two values for intensity and distribution are then combined to obtain a predicted score for the assay. By comparing the measured score with a reference score obtained experimentally, the prognosis of endocrine therapy is possible.

US Pat. No. 5,281,517 (Bacus et al.), Which is incorporated herein by reference in its entirety, discloses a method for determining certain parameters, such as DNA, using image analysis means. Methods and devices are taught for selecting and analyzing cells or subpopulations of cell subjects. The cells are first marked for type by staining the cells with an alkaline phosphatase method that includes a monoclonal antibody specific for a protein in at least one of the cell cytoplasm or a protein on the cell membrane. A second staining of DNA in the nucleus is performed by the Feulgen method, which destroys the cytoplasm of the cells. After staining and marking the cells, using image analysis means for visual parameters such as colored DNA or colored antigen,
The gating may be performed on the measured sub-population. Next, the selected cells are examined by digital image processing, and the true measured values of DNA (unit: picogram)
And other parameters. Actual measured values of the parameters are generated and obtained.

US Pat. No. 5,428,690 (Bacus et al.), Which is hereby incorporated by reference in its entirety, describes an automated assay for biological analytes placed on microscope slides. Apparatus and method for teaching. The apparatus includes an interactive optical subsystem for viewing a biological sample on a slide and generating an interactive video signal corresponding to a viewed image. The automated optics subsystem includes a single high power microscope objective for scanning a rack of slides, some of which are pre-identified for assays in interactive optics. This system also includes a processor that processes interactive and automatic video signals for the two optical subsystems. The processor receives the automatic video signal and performs a biological assay function upon receipt. Also disclosed are methods and apparatus for marking points on a microscope slide for subsequent analysis and for associating an analysis function with each marked point.

US Pat. No. 5,252,487 (Bacus et al.), Which is incorporated herein by reference in its entirety, measures the amount of oncogene protein product copy in a cell. Apparatus and method for the same. The device includes an optical conversion module for measuring the amount of optically enhanced DNA in a cell sample. A subsystem for measuring the amount of the optically enhanced oncogene protein product is coupled to the DNA measurement means. A subsystem for comparing the measurement of the amount of DNA with the measurement of the amount of the oncogene protein product provides a measurement of the copy of the oncogene protein product, which is supplied to an output device to provide a cell sample of the cell sample. An output is generated that is representative of the amount of the oncogene protein product in.

US Pat. No. 5,288,477 (Bacus et al.), Which is hereby incorporated by reference in its entirety, describes the efficacy of chemotherapy using monoclonal antibodies and ligand molecules. Is taught. The putative anticancer agent is HER-
It has binding specificity for oncogene receptor molecules on the membrane of cancer cells such as 2 / neu. When a putative drug binds to an oncogene receptor, the receptor:
From the membrane, it translocates to the cytoplasm or perinuclear of the cancer cells, with a transient increase in the total cellular content of the receptor, resulting in terminal cell differentiation. The efficacy of a drug in vivo can be measured in vitro by treating cancer cells that have undergone biopsy with this drug and then testing the cells for terminal cell differentiation. Such confirmations include morphological changes, decreased cell growth, or the production of chemicals associated with the mature phenotype. In addition, the treated cells may be examined with an immunohistochemical specific for the oncogene receptor to determine the translocation of the receptor from the membrane to the cytoplasm or perinuclear. Quantification of receptor levels in treated cells by measuring optical density after staining can be used to measure not only migration but also a transient increase in the total cellular content of the receptor.

US Pat. No. 5,134,662 (Bacus et al.), Which is hereby incorporated by reference in its entirety, describes an automated classification of cells and other microscopic specimens. Teach methods and apparatus used in the method. The device provides an operator-device interactive classification system for cell analysis, an alternative approach for different cells, cytoplasms and cell populations, and a compact adjustable assembly that can be operated to obtain enhanced image or color separation and analysis. You.

US Pat. No. 5,473,706 (Bacus et al.), Which is hereby incorporated by reference in its entirety, describes an automated assay for biological analytes placed on microscope slides. Devices and methods are taught. The apparatus includes an interactive optical subsystem for viewing a biological sample on a slide and for generating an interactive video signal corresponding to the viewed image. The automated optics subsystem includes a single high power microscope objective for scanning a rack of slides, some of which are pre-identified for assays in the interactive optics subsystem. The system also includes a processor that processes interactive and automatic video signals from the two optical subsystems. The processor receives the automatic video signal and performs a biological assay function upon receipt.

US Pat. No. 5,526,258 (Bacus et al.), Which is incorporated herein by reference in its entirety, discloses a method for analyzing a cellular subject in a cell sample to determine the actual cancer or its cancer. Devices and methods for diagnosing and treating suspected cancer are taught. The image of the cell sample is first digitized, and morphological attributes, including the area and DNA mass of the cell object, are automatically measured from the digitized image. The measured attribute is compared to a pre-determined attribute value range to select a particular cell target that has a value in the cancer analysis. After selecting the cell objects, an image is displayed to the operator, and an indication of the selection is displayed for each selected cell object. Next, the operator reviews the automatically selected cell objects in view of the benefits of the measured cell object attribute values and accepts or changes the automatic selection of cell objects. In a preferred embodiment, each selected cell object is assigned to one of six types. The indication of selection consists of an indication of the type in which the relevant cell object is located. In addition, the measured DNA mass of the fragment to be identified in the tissue section sample is increased to indicate the DNA mass of the whole cell, from which the fragment is cut.

US Pat. No. 5,546,323 (Bacus et al.), Which is incorporated herein by reference in its entirety, discloses an automated imager, preferably using polyploid nuclear DNA content. An apparatus and method for measuring the thickness of a tissue section by an analysis system is taught. The measured thickness is then used in the analysis of cellular objects in a test cell sample to monitor the diagnosis and treatment of actual or suspected cancer, or the variation of the nominal thickness in a microtome setting. First, an image of a measurement substance such as a rat liver tissue section having a known cell target attribute is digitized, and morphological attributes including the cell target area and DNA mass are automatically measured from the digital image.
The measured attribute is compared with an attribute value range determined in advance for selecting a specific cell object. After selecting the cell object, the operator automatically reviews the selected cell object and accepts or changes the measured cell object attribute value. In a preferred embodiment, each selected cell subject is assigned to one of three types corresponding to diploid, tetraploid and octaploid cell morphologies, and the identified cells in a rat liver tissue section sample The measured DNA mass of the fragment of interest may be corrected. Next, the mass of the selected cell target, for example, DNA, of the test substance is graphically displayed in a histogram, and the thickness of the rat liver tissue section can be measured based on the distribution.

US Pat. No. 5,018,209 (Bacus et al.), Which is hereby incorporated by reference in its entirety, describes a cell or cell subject for certain parameters such as DNA, estrogen, etc. After selecting and analyzing the subpopulations, a method and apparatus for measuring the selected cells is taught. After observing the field of view of the cells in real time, the observer selects cells having visual parameters such as colored DNA or colored antigen based on the morphological criteria and gates them to the subpopulation to be measured. The selected cells are examined by digital image processing to measure parameters such as the true measured value (unit: picogram) of DNA. Generate and obtain the quantity of the measured parameter.

US Pat. No. 4,998,284 (Bacus et al.), Which is hereby incorporated by reference in its entirety, describes a method for automatically classifying cells and other microscopic specimens. And apparatus. The device provides an operator-device interactive classification system for cell analysis, an alternative approach for different cells, cytoplasms and cell populations, and a compact adjustable assembly that can be operated to obtain enhanced image or color separation and analysis. You.

US Pat. No. 4,741,043 (Bacus et al.), Which is incorporated herein by reference in its entirety, discloses various cells, antigens or other substances obtained from the human body. It teaches a user interactive system for dynamically testing and evaluating. More specifically, DNA in a test cell is analyzed and quantified by image analysis using a pattern recognition method. The user prepares a unique slide or support on which the analyte and reference material or subject are simultaneously stained or otherwise image enhanced during the analysis.

The following conclusions can be drawn from the above considerations.

First, in situ analysis of neoplasms is likely to be useful in clinical applications, allowing for better diagnosis, prognosis and treatment.

Second, histostains, unique immunohistochemical markers and DNA ploidy stains (both conventional and immunostained) are all useful for this purpose.

Third, to date, cells have been labeled by combining multiple specific immunohistochemical stains with several histological stains and DNA ploidy stains, from which pathological data was extracted. It was not attempted or suggested. This is due to the fact that a system (imaging system) that can obtain an appropriate spatial resolution is widely used, but a system that can obtain an appropriate spatial resolution and an appropriate spectral decomposition is far less available.

However, such systems have been developed and have recently been used in a variety of applications, as described in further detail below.

A spectrometer is a device designed to receive light, separate it into its component wavelengths (disperse), and measure the light spectrum where the light intensity is a function of that wavelength. An imaging spectrometer collects the incident light from the scene, and each pixel (ie, pixel)
Is a spectrometer that measures the spectrum of

Spectroscopy is a well-known analytical tool that has been used for decades in science and industry to characterize materials and processes based on the spectral signature of chemical components.
The physical basis of spectroscopy is the interaction of light with matter. Traditionally, spectroscopy measures the intensity of light emitted, scattered or reflected from, or transmitted from, a sample in relation to wavelength with high spectral resolution but no spatial information.

On the other hand, spectral imaging is a combination of high-resolution spectroscopic analysis and high-resolution imaging (ie, spatial information). Most of the work described so far obtains high spatial resolution information from biological samples and provides limited spectral information when performing high spatial resolution imaging with one or several separate bandpass filters, for example. [Andersson-En]
gels et al. (1990) Proceedings of SPIE-Bioim
aging and Two-Dimensional Spectrosco
py (SPIE-Procedure for bioimaging and two-dimensional spectroscopy), 1205
179-189], or high spectral resolution (eg full spectrum)
And the spatial resolution is limited to a few points in the sample or averaged over the sample [see, for example, US Pat. No. 4,930,516 (Alfan
o))].

Conceptually, the spectral imaging system comprises (i) a measurement system and (i)
i) analysis software. The measurement system includes all of the optics, electronics, and method of illuminating the sample (eg, light source selection), measurement mode (eg, fluorescence or transmission), and calibration that is optimal to extract the desired result from the measurement. Analysis software includes all of the software and mathematical algorithms needed to analyze and display important results in a meaningful way.

Spectral imaging has been used for decades in the study of Earth and other planets in the area of remote sensing, which provides important information by identifying the characteristic spectral absorption features from which it is derived. However, remote sensing spectral imaging systems (eg, Landsat, AV
The high cost, size, and configuration of IRIS) limits their use to those arising from airplanes and satellites [Maymon and Neck (1988) P
rosedings of SPIE-Recent Advances i
n Sensors, Radiometry and Data Processes
sing for Remote Sensing (Procedure for SPIE-Recent Advances in Sensors, Radiometry and Data Processing for Remote Sensing),
924, pp. 10-22; Dozier (1988) Proceedings.
of SPIE-Recent Advances in Sensors, R
adimetry and Data Processing for Re
move Sensing (Procedure of SPIE-remote sensing sensor,
Recent Advances in Radiology and Data Processing), 924, pp. 23-30].

There are three basic types of spectral dispersion methods that can be used in spectral bioimaging systems: (i) spectral gratings or prisms, (ii) spectral filters, and (iii) buffered spectroscopy. As will be explained below, the latter is optimal for implementing the method of the invention. As will be appreciated by those skilled in the art, grating, prism and filter based spectral bioimaging systems may also be useful for certain applications.

A grating or prism (ie, monochromator) based system, also known as a slit imaging spectrometer (eg, DILOR system) [SPIE Conference E, opened in Barcelona, Spain]
European Medical Optics Week, BIOS Eur
ope 1995, see Valisa et al. (September 1995)]
Only one axis (spatial axis) of a CCD (Charge Coupled Device) array detector forms true image data, while the second (spectral) axis defines the intensity of light dispersed by the grating or prism as wavelength. Used for sampling in relation to The system also includes a slit in the first focal plane to limit the field of view at a given time to a line of pixels. Thus, a full image can only be obtained after scanning the grating (or prism) or incident light in a direction parallel to the spectral axis of the CCD in a manner known in the literature as line scanning. Since the two-dimensional image cannot be visualized before the entire measurement is completed, it is not possible to select an intended desired region from the field of view and / or optimize the system focus, exposure time, etc. before the measurement. Grating and prism-based spectral imagers have been used for remote sensing applications. This is because an airplane (or satellite) flying over the earth's surface results in a system with a natural line scanning mechanism.

Further, slit imaging spectrometers do not measure most of the pixels in one frame at a given time, even if the optics located in front of the instrument actually collect the incident light from all of them at the same time So there is a big drawback. as a result,
The measurement time required to obtain the required information with a constant S / N ratio is relatively long or the S / N ratio (sensitivity) is substantially reduced for a fixed measurement time. In addition, slit-type spectral imagers require line scanning to gather the necessary information for the entire scene. For this reason, the obtained result may be inaccurate.

[0117] Filter-based spectral dispersion methods can be further classified into discrete filters and tunable filters. In these types of imaging spectrometers, the spectral image is obtained by continuously inserting narrow band filters in the optical path, or by combining these bands with an acousto-optic tunable filter (AOTF) or a liquid crystal tunable filter (LCTF). By using electronic scanning, it is formed by simultaneously filtering all pixel radiation of the scene at different wavelengths at once (see description below). Most of the radiation is eliminated in a fixed amount of time in a similar manner to a slit imaging spectrometer with a grating or prism as described above, using a filter-based spectral dispersion method. In fact, it is possible to measure the entire image at a particular wavelength since all photons outside the instantaneous wavelength being measured are rejected and do not reach the CCD.

Tunable filters such as AOTF and LCTF have no moving parts and can be tuned to any particular wavelength within the spectral range of the device being implemented. One advantage of using tunable filters as a dispersion method for spectral imaging is that random wavelength access is possible (ie, the ability to measure image intensity at multiple wavelengths in the desired sequence without using a filter wheel). It is. However, for AOTF and LCTF, (i) the spectral range is limited (typically λ _max = 2λ _min ) and all other radiation outside this spectral range must be blocked, (ii) ) Temperature sensitive, (iii) poor transmittance, (iv) polarization sensitive, (v)
In the case of AOTF, the disadvantage is that the image shifts during wavelength scanning, after which careful and complex alignment is required.

All of these types of filters and tunable filter-based systems
Neither application has been successfully and widely used in spectral imaging for many years. This is due to limited spectral resolution, low sensitivity, inability to use easily, and the complexity of software algorithms used to interpret and display the data.

A method and apparatus used for spectral analysis of images having the above advantages is described in US patent application Ser. No. 08 / 392,019 (Cabib et al .; Feb. 2, 1995).
(Filed 1 day) (currently US Pat. No. 5,539,51, issued Jul. 23, 1996)
No. 7), which is incorporated herein by reference in its entirety. The purpose of this is to make better use of all the information obtained from the focused incident light of the image and to substantially reduce the required frame time, compared to a conventional slit or filter type imaging spectrometer, And / or substantially increasing the signal-to-noise ratio used in spectral analysis of images, and does not include line scanning. According to the present invention, there is provided a method of analyzing an optical image of a scene and measuring the spectral intensity of each pixel by collecting incident light from the scene, comprising the following steps: Passing the light through an interferometer that outputs modulated light corresponding to a predetermined linear combination of the spectral intensities of the light emitted from each pixel; placing the light output from the interferometer on a detector array Focusing; scanning the optical path difference (OPD) occurring in the interferometer for every pixel independently and simultaneously; and processing the output of the detector array (processing the interferogram of every pixel separately). And determine the spectral intensity of each pixel. The method can be implemented by utilizing various types of interferometers. In these interferometers, the interferogram is created by changing the OPD and moving the entire interferometer, elements within the interferometer, or the angle of incidence of the incident radiation. In all these cases, once the scanner has completed one scan of the interferometer, the interferogram for all pixels in the scene will be complete.

An apparatus having the above characteristics is different from a conventional slit type and filter type imaging spectrometer in that an interferometer as described above is used. Thus, by not limiting the collected energy with apertures or slits, or limiting the incident wavelength with narrow band interference or tunable filters, the overall throughput of the system is substantially increased. Therefore, the interferometer system substantially reduces the measurement time by better utilizing all the information obtained from the incident light of the analyzed scene,
And / or substantially increase the S / N ratio (ie, sensitivity). The sensitivity advantages of interferometer spectroscopy over filter and grating or prism methods are:
It is known in the art as a multiplex or Fellgett advantage [Chamberlain (1979) The principals of.
interferometric spectroscopy (principle of interferometric spectroscopy), John Wiley and Sons, pages 16-18 and 263].

For example, John B. Wellman (1987), Imaging Sp.
electrometers for Terrestrial and Plan
Consider the "whiskroom" design described in "Early Remote Sensing" (Imaging Spectrometer for Remote Sensing of Earth and Planets), SPIE Proceedings, Vol. 750, p. Let n be the number of detectors in the linear array, mxm be the number of pixels in the frame, and T be the frame time. The total time taken for each pixel in one frame summed over all the detectors of the array is nT / m ^2. By using the same size array and the same frame rate as the method according to the invention described in US Pat. No. 5,539,517, the total time taken for all detectors for a particular pixel is the same, ie , NT / m ² . However, in conventional grating or prism methods, the energy seen by all detectors at any given time is on the order of 1 / n of the total energy. This means that the modulation function is an oscillation function (eg, Michelson).
Or a similar periodic function (e.g., low fineness Air with Fabry-Perot)
Since the y-function (average over the large OPD range is 50%)), the wavelength resolution is less than that of the method according to the invention described in US patent application Ser. This is due to 1 / n of the range. Textbook on interferometry [eg, Chamberlain (1979) The prince]
chips of interferometric spectrosco
py, John Wiley and Sons, pages 16-18 and 263.
Based on the standard treatment of the Fellgett Advantage (or Multiplex Advantage) described in the paragraph [1], the device according to the invention can be used in the case of a noise limitation in which the noise level is independent of the signal (system or background noise is limited). ), The measured S / N ratio is improved by n ^0.5 times, and if the limitation is due to signal photon noise, in the spectral range at the narrow peak wavelength,
It can be shown that the signal at a particular wavelength is improved by the square root of the ratio to the average signal. Thus, according to the invention described in U.S. Pat. No. 5,539,517, all required OPDs are converted to the scene to obtain all the information needed to reconstruct the spectrum. Scanning for all pixels simultaneously results in simultaneous acquisition of spectral information with imaging information.
In the present invention, a number of different optical configurations are available, including optical instruments for remote sensing, microscopes for laboratory analysis, fundus cameras for retinal imaging, optical fibers and endoscopy for industrial monitoring and medical imaging, diagnosis, treatment, etc. A mirror can be used.

Continuation applications (US patent application Ser. No. 08 / 571,047 (Cabib et al .; 1995)
The aim of the present application is to provide spectral imaging methods for biological research, medical diagnostics and therapeutics. . These methods rely on high spatial and high spectral resolution light transmission, reflection, scattering and fluorescence emission techniques to achieve spatial organization (ie, distribution).
And quantifying the natural components and structures of cells and tissues, organelles and components to be administered such as labeled probes (ie, fluorescent probes) and drugs. In US patent application Ser. No. 08 / 571,047, U.S. Pat. The use of the spectral imaging device described in U.S. Pat. No. 5,39,517 has been demonstrated, as well as additional biological and medical applications.

[0124] Spectral bioimaging systems can be useful in all applications where there are subtle spectral differences between chemical components (their spatial distribution and tissue in the image is what is intended). This measurement is described in US Pat. No. 5,539,5.
Substantially any optical system attached to the system of No. 17, such as a vertical or inverted microscope, a fluorescent microscope, a macro lens, an endoscope and a fundus camera,
Can also be used. In addition, any standard experimental method may be used, including light transmission (brightfield and darkfield), autofluorescence and dosing probe fluorescence.

The fluorescence measurements are based on the assumption that the emission spectrum is within the spectral range of the system sensitivity, and can be any standard filter cube (consisting of a barrier filter, an excitation filter and a dichroic mirror), or a spatial filter. It can be implemented with any customized filter cube for the application. Spectral bioimaging can also be used in conjunction with any of the standard spatial filtration methods (eg, darkfield and phase contrast), and even with a polarizing microscope. Of course, the effect on spectral information when using such a method must be understood in order to correctly interpret the measured spectral image.

US patent application Ser. No. 08 / 824,234, filed Mar. 25, 1997, which is incorporated herein by reference in its entirety, describes a cell (eg, a cancer cell). It teaches a method for morphometric classification (eg, detection, grading) with automatic and / or semi-automatic spectral decomposition of neoplasms, configured to classify objectively rather than subjectively. According to the method disclosed herein, (a) providing a sample containing at least a portion of at least one cell to be subjected to spectral imaging; (b) via an optical device optically connected to an imaging spectrometer. Inspecting the sample to obtain a spectrum for each pixel of the sample; (c) classifying each of the pixels into a classification group according to the spectrum of the pixel; (d) analyzing the classification group of pixels. The cells of the sample are classified by cell type.
This method can be applied, for example, to counterstained breast carcinoma.

The combination of spectral imaging, in situ counterstaining and in situ immunostaining is the first suggested and attempted here. Such a combination allows simultaneous detection of a plurality of immunohistochemical stains, tissue stains, and DNA ploidy stains, which were not possible with the prior art imaging methods. Therefore, such a combination would revolutionize the field of cytopathology by combining data extracted simultaneously from various histostains, DNA ploidy stains and specific immunohistochemical stains. It is.

Therefore, a plurality of immunohistochemical stains, tissue stains, and DNAs were obtained by using a spectral imaging method having high spatial resolution and high spectral resolution.
The need for a method for simultaneous in situ analysis of ploidy stains is widely recognized,
And it would be very advantageous to have such a method.

DISCLOSURE OF THE INVENTION According to the present invention, there is provided a method for in situ analysis of a biological sample, wherein (a) the biological sample is prepared by a method in which the first stain is a first immunohistochemical stain, a first tissue stain, The second DNA stain is selected from the group consisting of a first DNA ploidy stain, and the second stain comprises a second immunohistochemical stain, a second tissue stain, and a second DNA ploidy stain. A step of staining with N kinds of staining agents selected from the group, wherein N is an integer greater than 3, and (i) the first staining agent is the first immunohistochemical staining agent Wherein the second stain is the second tissue stain or the second DNA ploidy stain, and (ii) wherein the first stain is the first tissue stain. Is
The second stain is the second immunohistochemical stain or the second DNA ploidy stain, while (iii) the first stain is the first DNA ploidy stain Is a step in which the second stain is the second immunohistochemical stain or the second tissue stain, and (b) collecting spectral data from the biological sample using a spectral data collection device Selecting the spectral data acquisition device and the N stains such that spectral components associated with each of the N stains can be acquired.

According to still further features in preferred embodiments of the invention described below, a method for in situ analysis of a biological sample, the method comprising: (a) combining a biological sample with a first immunohistochemical stain. A first tissue stain and a first DNA ploidy stain, wherein the second stain is a second immunohistochemical stain, a second tissue stain, A step of staining with a plurality of stains selected from the group consisting of DNA ploidy stains, provided that (i) the first stain is the first immunohistochemical stain In this case, the second stain is the second tissue stain or the second DNA ploidy stain, and (ii) when the first stain is the first tissue stain, The second stain is the second immunohistochemical stain or the second DN. A ploidy stain, while (iii) the first stain is the first D
In the case of an NA ploidy stain, the second stain is the second immunohistochemical stain or the second tissue stain, and (b) using a spectral data collection device. A step of collecting spectral data from a biological sample, wherein the spectral data collecting device and the plurality of types of stains are selected such that spectral components associated with each of the plurality of types of stains can be collected, Are provided.

According to still further features in preferred embodiments of the invention described below, a method for in situ analysis of a biological sample, wherein (a) the biological sample is stained with at least four different immunohistochemical stains. And (b) collecting spectral data from the biological sample using a spectral data collecting device, wherein the spectral data collecting device and the at least four types of immunohistochemical stains are combined with the at least four types of immunohistochemical stains. Selecting such that spectral components associated with each of the immunohistochemical stains can be collected.

According to still further features in preferred embodiments of the invention described below, a method for in situ analysis of a biological sample, wherein (a) the biological sample is at least one staining agent is an immunohistochemical staining agent. Staining with at least three stains, wherein the at least one additional stain is a tissue stain or a DNA ploidy stain;
(B) collecting spectral data from the biological sample using a spectral data collecting device, wherein the spectral data collecting device and the at least three dyes are associated with each of the at least three dyes; Selecting the collected spectral components so that they can be collected.

According to still further features in preferred embodiments of the present invention described below, a method for in situ analysis of a biological sample, the method comprising: (a) transferring the biological sample, wherein the first stain is an immunohistochemical stain; A step of staining with at least three types of stains, wherein the second stain is a tissue stain and the third stain is a DNA ploidy stain, and (b) using the spectral data collection device to Collecting spectral data, selecting the spectral data collection device and the at least three dyes such that spectral components specifically associated with each of the at least three dyes can be collected. And a method comprising:

According to still further features in the described preferred embodiments the invention comprises at least four different immunohistochemical stains, each immunohistochemical stain staining a respective cytological marker. An immunohistochemical composition wherein each immunohistochemical stain is capable of being individually detected in the presence of all other immunohistochemical stains using a spectral data collection device. Is provided.

According to still further features in preferred embodiments of the invention described below, the first immunohistochemical stain and the second immunohistochemical stain each independently activate a primary antibody and a signal amplification mechanism. Including.

According to further features in preferred embodiments of the invention described below, the signal amplification mechanism comprises a second antibody capable of binding a constant region of the first antibody, and a biotin conjugated to the first antibody. It is selected from the group consisting of avidin or streptavidin capable of binding and biotin capable of binding avidin or streptavidin conjugated to the primary antibody.

According to still further features in preferred embodiments of the invention described below, the secondary antibody, avidin, streptavidin and biotin are each independently labeled with a detectable moiety.

According to further features in preferred embodiments of the invention described below, the detectable component is a fluorescent dye.

According to further features in preferred embodiments of the invention described below,
The fluorescent dye is fluorescein, rhodamine, Texas Red, Cy2, Cy3,
Cy5, Cy0, Cy0.5, Cy1, Cy1.5, Cy3.5, Cy7, VE
CTOR Red, ELF ™ (enzyme-labeled fluorescein), FluorX,
Calcein, Calcein-AM, CRYPTOFLUOR ™ (orange (
42kDa), Tangerine (35kDa), Gold (31kDa), Red
(42 kDa), Crimson (40 kDa)), BHMP, BHDMAP, Br
-Oregon, Lucifer Yellow, Alexa dyes, N- [6- (7-nitrobe
2-oxa-1,3-diazol-4-yl) amino] caproyl] (N
BD), BODIPY ™, boron dipyrromethene difluoride, oregong
Lean, MITOTRACKER ™ Red, DiOC₇(3), DiIC ₁₈ , Phycoerythrin, phycobiliprotein BPE (240 kDa) RPE
(240 kDa) CPC (264 kDa) APC (104 kDa), spectrum
Blue, Spectrum Aqua, Spectrum Green, Spectrum Gold, Spec
Tolu orange, spectral red, NADH, NADPH, FAD, infrared (IR
) Dyes, cyclic GDP ribose (cGDPR), chalcofluor white, fir
And tryptophan.

According to further features in preferred embodiments of the invention described below, the detectable component is a non-fluorescent dye.

According to further features in preferred embodiments of the invention described below, the non-fluorescent dye comprises the group consisting of an alkaline phosphatase substrate, a horseradish peroxidase substrate, a glucose oxidase substrate, and a β-galactosidase substrate. Is selected from

According to further features in preferred embodiments of the invention described below, the detectable component is an enzyme that catalyzes a colorimetric reaction of a substrate having a substantially insoluble color reaction product.

According to further features in preferred embodiments of the invention described below, the enzyme is selected from the group consisting of alkaline phosphatase, horseradish peroxidase, β-galactosidase, and glucose oxidase. is there.

According to further features in preferred embodiments of the invention described below, the substrate is selected from the group consisting of an alkaline phosphatase substrate, a horseradish peroxidase substrate, a β-galactosidase substrate, and a glucose oxidase substrate. It was done.

According to still further features in preferred embodiments of the invention described below, the detectable component is capable of emitting light or substantially directing a second reaction of a second substrate with a luminescent product. Is an enzyme that catalyzes a luminescence reaction of a substrate having a reaction product insoluble in water.

According to further features in preferred embodiments of the invention described below, the enzyme is selected from the group consisting of luciferase and aequorin.

[0147] According to further features in preferred embodiments of the invention described below, the first substrate and the second substrate, and each independently luciferin, and ATP, and Ca ^{+ +,} consisting of coelenterazine It was selected from a group.

According to still further features in preferred embodiments of the invention described below, the first tissue stain and the second tissue stain are each independently 4 ′, 6-diamidino-2-phenylindole. , Eosin, fluorescein isothiocyanate, Hoechst 33258, Hoechst 33342, propidium iodide, quinacrine, fluorescein phalloidin, resorufin, hematoxylin, orange G,
Light Green SF, Romanovsky-Giemsa, May-Grunwald, Blue counterstain, ethyl green, foilgen naphthol yellow S, Giemsa, methylene blue, methyl green, pyronin, naphthol yellow, neutral red, Papanicolaou stain, red counterstain C and Sirius Red.

According to further features in preferred embodiments of the invention described below, the first DNA ploidy stain and the second DNA ploidy stain are each independently chromomycin A3, DAPI, Flavin-Feulgen reaction, auramine O
-Feulgen reaction, ethidium bromide, propidium iodide, high-affinity DNA phosphor, green fluorescent protein fused to DNA binding protein, ACMA, quinacrine orange and acridine orange, Feulgen reagent,
Gallocyanine chrome alum, gallocyanine chrome alum and naphthol yellow S, methyl green pyronin Y and a thionine-Foilgen reagent.

According to further features in preferred embodiments of the invention described below, the spectral data acquisition device is an interferometer-based spectral data acquisition device, a filter-based spectrum data acquisition device, and a dispersive element-based spectrum data acquisition device. Selected from the group consisting of

According to further features in preferred embodiments of the invention described below, the first immunohistochemical stain and the second immunohistochemical stain each independently comprise a primary antibody.

According to further features in preferred embodiments of the invention described below, the primary antibody is an anti-estrogen receptor antibody, an anti-progesterone receptor antibody, an anti-p53 antibody, an anti-Her-2 / neu antibody, an anti-EGFR antibody , Anti-cathepsin D
Antibody, anti-Bcl-2 antibody, anti-E-cadherin antibody, anti-CA125 antibody, anti-CA15
-3 antibody, anti-CA19-9 antibody, anti-c-erbB-2 antibody, anti-P-glycoprotein antibody, anti-CEA antibody, anti-retinoblastoma protein antibody, anti-ras tumor protein antibody, anti-Lewis X antibody, anti-Ki -67 antibody, anti-PCNA antibody, anti-CD3 antibody, anti-CD4 antibody, anti-CD5 antibody, anti-CD7 antibody, anti-CD8 antibody, anti-CD9 / p24 antibody, anti-CD10 antibody, anti-CD11c antibody, anti-CD13 antibody, anti-CD14 antibody, Anti-CD15 antibody, anti-CD19 antibody, anti-CD20 antibody, anti-CD22 antibody, anti-CD23
Antibody, anti-CD30 antibody, anti-CD31 antibody, anti-CD33 antibody, anti-CD34 antibody, anti-CD35 antibody, anti-CD38 antibody, anti-CD41 antibody, anti-LCA / CD45 antibody, anti-CD45RO antibody, anti-CD45RA antibody, anti-CD39 antibody, anti-CD100 antibody,
Anti-CD95 / Fas antibody, anti-CD99 antibody, anti-CD106 antibody, anti-ubiquitin antibody, anti-CD71 antibody, anti-c-myc antibody, anti-cytokeratin antibody, anti-vimentin antibody, anti-HPV protein antibody, anti-kappa light chain antibody, anti-lambda L chain antibody, anti-melanosome antibody, anti-prostate specific antigen antibody, anti-S-100 antibody, anti-tau antigen antibody,
It is selected from the group consisting of an anti-fibrin antibody, an anti-keratin antibody and an anti-Tn antigen antibody.

According to the present invention, the present invention provides a method for simultaneous in situ analysis of a plurality of immunohistochemical stains, tissue stains, and DNA ploidy stains using a spectral imaging method of high spatial and spectral resolution. The disadvantages of the known arrangement can be overcome.

BRIEF DESCRIPTION OF THE DRAWINGS The invention will now be described, by way of example only, with reference to the accompanying drawings, in which: In the figures: FIG. 1 is a block diagram showing the main components of an imaging spectrometer constructed according to US patent application Ser. No. 08 / 392,019 (prior art). FIG. 2 shows a non-moving interferometer, or Sagnac interferometer, used in an imaging spectrometer according to US patent application Ser. No. 08 / 392,019 (prior art). 3A-C show mesophase FISH performed using two different probes conjugated to Texas Red and Rhodamine, FIG. 3A is an original image viewed through a microscope, and FIG. 3B is the same sample as FIG. 3A. About SPECTR
FIG. 3C shows an image after measurement and processing by the ACUBE ™ apparatus, and FIG. 3C is a fluorescence spectrum of the Texas Red phosphor and the Rhodamine phosphor. Figures 4A-C show SP for six different probes each labeled with a different fluorophore.
FIG. 4 is a mesophase FISH performed using an ECTRACUBE ™ apparatus,
A is the original image seen through the microscope, in which case the cells are D
FIG. 4B shows the results of counterstaining with the API.
FIG. 4 shows an image after measurement and processing by the ECTRACUBE (trademark) apparatus, and FIG.
C is the fluorescence spectrum of the six phosphors used for classification. FIGS. 5A to 5C show the results of hybridization between 24 kinds of chromosome paints and normal male chromosome spreads obtained using the SPECTRACUBE ™ apparatus, and FIG. 5A shows the results obtained by using the RGB algorithm. 5B and 5C are classification images obtained using the classification algorithm, FIGS. 5A and 5B show the original spread, and FIG. 5C is the karyotype. Shows the chromosome spread arranged as. FIG. 6 illustrates the definition of pseudo-RGB (red, green and blue) colors to enhance selected spectral ranges. The intensity of each pseudo color is calculated by multiplying the area under the curve by one of the curves and then integrating this. FIG. 7 shows SPECTRACUBE (breast cancer samples) stained with six single stains.
2) Histostains (hematoxylin and eosin) and 4 immunohistochemical stains (DAB, Fast Red, AEC and BCI) as measured using
2 shows an unnormalized spectrum of (P / NBT). Peak wavelengths are shown on the right. 8A-E show tissue stain hematoxylin and immunohistochemical stain anti-ER-DA.
8B is a diagram showing an image of a breast cancer sample previously known to be ER (+) / PR (+) co-stained with B, FIG. 8A shows an RGB image of the sample, and FIGS.
D is the hematoxylin spectral component, DAB spectral component and AE
FIG. 8E represents a binarized image of the C spectral components, while FIG. 8E represents a classified overlay image with the spectral components highlighted in red, green and blue,
These were all measured using a SPECTRACUBE ™ instrument. FIGS. 9A-E show the tissue stain hematoxylin and the immunohistochemical stain anti-PR-AE.
9C is a diagram showing an image of a breast cancer sample previously known to be ER (+) / PR (+) co-stained with C, FIG. 9A shows an RGB image of the sample, and FIGS.
D is the hematoxylin spectral component, DAB spectral component and AE
FIG. 9E represents a binarized image of the C spectral components, while FIG. 9E represents a classified overlay image with the spectral components highlighted in red, green and blue,
These were all measured using a SPECTRACUBE ™ instrument. FIGS. 10A-E are images showing breast cancer samples previously known to be ER (+) / PR (+) co-stained with the tissue stain hematoxylin and the immunohistochemical stain anti-PR Fast Red, FIG. 10A shows an RGB image of the sample,
10B to 10D show binarized images of a hematoxylin spectral component, a DAB spectral component, and a fast red spectral component, respectively, while FIG.
0E represents a classified overlay image in which the above spectral components are emphasized in red, green and blue, respectively, all of which were measured using a SPECTRACUBE ™ instrument. FIGS. 11A-E show the tissue stain hematoxylin and the immunohistochemical stain anti-ER-D.
FIG. 11B is an image of a breast cancer sample previously known to be ER (+) / PR (+) co-stained with AB and anti-PR-Fast Red, and FIG.
11B-11D show the binarized images of the hematoxylin, DAB, and fast red spectral components, respectively, while FIG. 11E shows the spectral components highlighted in red, green, and blue, respectively. The classified overlay image is shown, and the area co-stained with anti-ER-DAB and anti-PR-fast red is shown in yellow. The above images are all S
This was measured using a SPECTRACUBE (trademark) apparatus. FIGS. 12A-F show ER (+) previously co-stained with the histostains hematoxylin and eosin and the immunohistochemical stains anti-ER-DAB and anti-PR-fast red.
FIG. 12 shows an image of a breast cancer sample known to be / PR (+).
A represents an RGB image of a sample, and FIGS. 12B to 12E represent binarized images of a hematoxylin spectral component, an eosin spectral component, a DAB spectral component, and a fast red spectral component, respectively, while FIG. Represents a classified overlay image with blue, purple, green and red, respectively. The above images were all measured using a SPECTRACUBE ™ instrument. Figure 13 shows SPECTRACU from cervical cancer samples stained with five single stains.
FIG. 3 is an unnormalized spectrum of five tissue stains (Harris Hematoxylin, Eosin, Orange G, Light Green SF and Bismarck Brown Y) measured using a BE ™ instrument. Peak wavelengths are shown on the right. FIGS. 14A-G are co-stained with Harris Hematoxylin, Eosin, Orange G, Light Green SF and Bismarck Brown Y, which are tissue stains, together forming what is known in the art as Papanicolaou stain. FIG. 14A shows an image of a cervical cancer sample obtained, FIG. 14A shows an RGB image of the sample, and FIG.
4B to 14F represent binarized images of the Harris hematoxylin spectral component, the eosin spectral component, the orange G spectral component, the light green SF spectral component, and the Bismarck Brown Y spectral component, respectively.
E represents a classification overlay image, in which the above spectral components are emphasized in blue, pink, orange, green and gray, respectively, and a combination of them is used to form a super-color classification overlay image. All the above images are SP
It was measured using an ECTRACUBE ™ instrument. BEST MODE FOR CARRYING OUT THE INVENTION The present invention relates to a method for simultaneous in situ analysis of a plurality of immunohistochemical stains, tissue stains and / or DNA ploidy stains that can be used for pathological examination of cells. That is, the present invention can be used to provide pathologists with cumulative information about a biological sample that has been examined, to aid in decision making.

The principles and operation of a method according to the present invention may be better understood with reference to the drawings and accompanying descriptions.

Before describing at least one embodiment of the present invention in detail, the present invention is not limited in its application to the details of construction and the arrangement of components set forth in the following description or drawings. You have to understand. The invention is capable of other embodiments and of being practiced and carried out in various ways. Further, the terms and technical terms used herein are used for the purpose of explanation, and do not limit the present invention.

Recent studies have shown that optical mixtures of several absorption dyes can be analyzed in certain cases. If the absorption bands of the individual stains do not overlap, the analysis is simple and easy. The wavelength is selected such that each component sequentially exhibits a suitable absorption for the measurement and the other components exhibit negligible absorption. That is, each dye in the mixture can be measured by performing one attenuation measurement at each selected wavelength.
This method has been successfully used by Bacus et al. This method has the advantage that (i) the biochemical composition of the cell or tissue is known; (ii) the use of a narrow band filter allows easy imaging. Unfortunately, with current cytochemistry and histochemistry, the number of such non-overlapping dyes in the spectrum is very limited.

Thus, according to one aspect of the present invention, there is provided a method for in situ analysis of a biological sample. This method can be performed by performing the following steps. In the first step, the biological sample is prepared by using the first stain as the first immunohistochemical stain, the first histostain or the first DNA ploidy stain, and the second stain as the second immunohistochemical stain. , A second tissue stain or a second DNA ploidy stain, with N (ie, multiple) stains.

Here, N is an integer larger than 3 (for example, a number of 4, 5, 6, 7, 8, 9, 10, or more, for example, an integer of 11 to 50 (including 11 and 50)). And (i) when the first stain is the first immunohistochemical stain, the second stain is the second tissue stain or the second DNA ploidy stain. Yes;
(Ii) when the first stain is the first tissue stain, the second stain is the second immunohistochemical stain or the second DNA ploidy stain; (Iii) When the first stain is the first DNA ploidy stain, the second stain is the second immunohistochemical stain or the second tissue stain.

In the second step of the method of the present invention, spectral data is collected from the biological sample using a spectral data collecting device. At this time, the spectral data collection device and the N types of stains are selected so that spectral components related to each of the N types of stains can be collected.

According to another aspect of the present invention, another method for in situ analysis of a biological sample is provided. This method is performed by performing the following steps. That is, in the first step, the biological sample is stained with at least four different immunohistochemical stains. In the second step of the method, spectral data is collected from the biological sample using a spectral data collecting device. At this time, the spectral data collecting device and the at least four types of immunohistochemical stains are selected so as to collect spectral components associated with each of the at least four types of immunohistochemical stains.

According to yet another aspect of the present invention, there is provided yet another method of in situ analysis of a biological sample. This method is performed by performing the following steps. That is, in the first step, the biological sample is subjected to at least three types of stains, wherein at least one stain is an immunohistochemical stain and at least one further stain is a tissue stain or a DNA ploidy stain. Stain with In the second step of the method, spectral data is collected from the biological sample using a spectral data collecting device.
At this time, the spectral data collection device and the at least three dyes are selected so that spectral components associated with each of the at least three dyes can be collected.

According to yet another aspect of the present invention, there is provided yet another method of in situ analysis of a biological sample. This method is performed by performing the following steps. That is, in the first step, the biological sample, the first stain is an immunohistochemical stain,
Stain with at least three stains, the second stain being a tissue stain and the third stain being a DNA ploidy stain. In the second step of the method, spectral data is collected from the biological sample using a spectral data collecting device. At this time, the spectral data collecting device and the at least three types of stains are selected so that spectral components specifically associated with each of the at least three types of stains can be collected.

According to a further aspect of the present invention, at least four different immunohistochemical stains are included in the mixture, each immunohistochemical stain for staining a respective cytological marker in a biological sample. Wherein each immunohistochemical stain is capable of being individually detected in the presence of all other immunohistochemical stains using a spectral data collection device. Provided.

As used herein and in the claims, the term “in situ” or “
"In situ analysis" refers to the analysis of a cell or tissue component that is located and preferably fixed at a natural location or location within a cell or tissue.

As used herein and in the claims, the term “biological sample” refers to an animal (
(Mammalian, especially human). This sample contains healthy tissue,
It may be diseased tissue or tissue suspected of being diseased tissue. The sample can be, for example, a biopsy taken during a surgical procedure. The sample can be collected by aspiration with a fine needle, scraping, or washing the cavity and collecting cells and tissues therefrom. The sample is
It may be both solid tumors and hematopoietic tumors, as well as adjacent healthy tissues. The sample may be a smear or piece of tissue of an individual cell.

As used herein and in the claims, the term “staining” refers to a process in which coloring occurs due to foreign material that has penetrated and / or interacted with the biological sample.

As used herein and in the claims, the term “stain” refers to fluorescent, luminescent and / or non-fluorescent colorants, and also refers to the reagents or substances used to effect coloration. means.

The term “immunohistochemical stain” as used herein and in the claims
Refers to colorants, reactions and related reagents that stain the biological sample under test directly or indirectly ("sandwich" reagents and / or enzymatic reactions) using primary antibodies that bind cytological markers. . Immunohistochemical stains are often referred to in the scientific literature as immunostains, immune cell stains, immunohistopathological stains, and the like.

As used herein and in the claims, the term “antibody” refers to a monoclonal or polyclonal immunoglobulin or a fragment of an immunoglobulin such as sFv (single-chain antigen binding protein), Fab1 or Fab2.

As used herein and in the claims, the term “tissue staining” refers to the type of protein (acidic or basic), DNA, RNA, lipid, cytoplasmic component, nuclear component, membrane component, etc. Refers to the colorants, reactions and / or related reagents used to stain cells and tissues in connection with. Histological stains are often referred to as counterstains, cytological stains, histopathological stains, and the like.

The term “DNA ploidy stain” as used herein and in the claims
Refers to a stain that stoichiometrically binds chromosomal components, including but not limited to DNA or histones. When an antibody such as an anti-histone antibody is involved, such stains are also known as DNA immunoploidy stains and the like.

The term “spectral data collection device” used in the present specification and claims means that a plurality (typically four or more) of each spatial element (pixel) of a test sample is used.
Means a device capable of detecting the light intensity associated with a separate spectral band of For example, a SPECTRACUBE ™ system optically connected to a microscope
Preferably, it plays a role as a spectrum data collecting device according to the present invention. However, any spectral imaging device, ie, a filter (eg,
Conventional acousto-optic tunable filter (AOTF) or liquid crystal tunable filter (LCTF) and dispersive elements (eg gratings or prisms)
For measuring and subsequently retrieving and analyzing the spectrum of light emitted at all points of an object placed in a field of view, including a system spectral imager, or other spectral data or multiband concentrator. Device to be stored in the storage device (for example, Sp
eicher R .; M. Ballard S .; G. FIG. And Ward C.A. D. (
1996) Karyotyping human chromasomes b
y combinatorial multifluor FISH, Natu
re Genetics, 12: 368-375) can be used to obtain the required spectral data. Also, a plurality of broadband filters as described in U.S. patent application Ser. No. 08 / 917,213, filed Aug. 25, 1997, which is incorporated herein by reference in its entirety. Devices that include fixed or tunable devices can also be used as spectral data acquisition devices according to the present invention. Accordingly, the scope of the present invention is not limited by the use of any particular type of spectral data acquisition device or by any particular type of spectral imaging device.

That is, the spectrum data acquisition device can be an interferometer-based spectrum data acquisition device, a filter (single or plural) -based spectrum data acquisition device, and a dispersive element-based spectrum data acquisition device.

The term “spectral component,” as used herein and in the claims,
Means a portion of the spectrum that is unique to a specific substance and can therefore be used as the substance's spectral signature to distinguish the substance from other substances.

According to one embodiment of the present invention, each of the immunohistochemical staining agents independently includes a primary antibody and a signal amplification mechanism. The signal amplification mechanism is, for example, a secondary antibody capable of binding the constant region of the primary antibody, avidin or streptavidin capable of binding biotin conjugated to the primary antibody, and avidin or streptavidin conjugated to the primary antibody Any available biotin can be used.

According to a preferred embodiment of the present invention, said secondary antibody, avidin, streptavidin or biotin is each independently a substantially insoluble color reaction product, a fluorescent dye (stain), Labeled with a detectable moiety, which can be an enzyme that induces a colorimetric reaction of a substrate with a luminescent or non-fluorescent dye. Examples of these choices are described below.

As the enzyme, for example, alkaline phosphatase, horseradish peroxidase, β-galactosidase and / or glucose oxidase can be used, and the substrates are alkaline phosphatase substrate, horseradish peroxidase substrate, β-galactosidase substrate or glucose oxidase substrate, respectively. Can be

In addition, the enzyme can emit light or have a second substrate having a luminescent product (including but not limited to luciferin and ATP or coelenterazine and Ca ⁺⁺ )
Can be an enzyme that catalyzes the luminescent reaction of a substrate having a substantially insoluble reaction product (including but not limited to luciferase and aequorin), which can lead to the second reaction of

Any tissue and DNA ploidy stains that can be used while practicing the methods of the invention are exemplified below.

According to a most preferred embodiment of the invention, each of the immunohistochemical stains used comprises a primary antibody.

According to yet another embodiment of the present invention, an external calibration is used to account for the inter-day variability seen when attempting staining. That is, it is preferable to try at least one calibration for each staining batch. For this purpose, a calibration material is used in which the biological sample and the calibration material are simultaneously stained with the same staining solution. First, using a spectrum data collection device, a stain calibration substance is analyzed so that adjustment of the spectrum data collection device or extraction of calibration data for calibration after algorithmic measurement can be performed. Only then can biological samples be analyzed more meaningfully, taking into account corrections based on calibration. Those skilled in the art will know how to devise a calibration algorithm that compensates for the inter-day variability seen when attempting staining.

The calibration substance can be an optical density reference substance. The calibrators can also include control cells that are co-stained together with the test biological sample.

The following is a list of various dyes or dyes that can be used to practice the method of the present invention. In addition, natural cell components with detectable spectral signatures that can be co-detected with stains using the methods of the invention are also listed.

It will be apparent to those skilled in the art that certain staining methods may interfere with certain other staining methods. Therefore, the types of dyes used and their order of application must be carefully considered. These considerations can be applied by those skilled in the art knowing how to stain.

Immunohistochemical stains for transmission microscopy : In principle, (i) can be conjugated or indirectly conjugated (eg, via conjugated avidin, streptavidin, biotin, secondary antibody) to the primary antibody, and (ii) 2.) Enzymes can be used that use soluble substrates to obtain insoluble products (precipitates). Such enzymes include, for example, HRP, AP, LacZ
And glucose oxidase.

Alkaline phosphatase (AP) substrates include AP-blue substrate (blue precipitate, catalog page 61 of Zymed), AP-orange substrate (orange precipitate, Zym)
ed), AP-red substrate (red precipitate, Zymed), 5-bromo, 4-
Chloro, 3-indolyl phosphate (BCIP substrate, blue-green precipitate), 5-bromo, 4-chloro, 3-indolyl phosphate / nitrobluetetrazolium / iodonitrotetrazolium (BCIP / INT substrate, tan precipitate, Bio
Meda), 5-bromo, 4-chloro, 3-indolylphosphate / nitrobluetetrazolium (BCIP / NBT substrate, blue / purple), 5-bromo, 4
-Chloro, 3-indolyl phosphate / nitroblue tetrazolium / iodonitrotetrazolium (BCIP / NBT / INT, brown precipitate, DAKO)
, Fast Red (Red), Magenta-Phos (Magenta), Naphthol AS-
BI phosphate (NABP) / fast red TR (red), naphthol A
S-BI phosphate (NABP) / new fuchsin (red), naphthol A
S-MX phosphate (NAMP) / new fuchsin (red), new fuchsin AP substrate (red), p-nitrophenyl phosphate (PNPP, yellow, water-soluble), VECTOR ™ black (black), VECTOR ™ blue (
Blue), VECTOR ™ red (red), Vega red (raspberry red)
And the like, but are not limited to these.

Horseradish peroxidase (HRP; sometimes abbreviated as “PO”) substrates include 2,2′-azino-di-3-ethylbenz-thiazoline sulfonate (ABTS, green, water-soluble), aminoethyl Carbazole, 3-amino, 9
-Ethylcarbazole AEC (3A9EC, red), α-naphtholpyronine (
Red), 4-chloro-1-naphthol (4C1N, blue, blue-black), 3,3
'-Diaminobenzidine tetrahydrochloride (DAB, brown), o-dianisidine (green), o-phenylenediamine (OPD, brown, water-soluble), TACS blue (blue), TACS red (red), 3,3', 5 , 5'-Tetramethylbenzidine (TMB, green or green / blue), TRUE BLUE (TM) (blue), VECTOR (TM) VIP (purple), VECTOR (TM) SG (Smoky Blue Gray) and Zymed Blue HRP Substrates (brilliant blue), but are not limited to these.

Glucose oxidase (GO) substrates include nitroblue tetrazolium (N
BT, purple precipitate), tetranitro blue tetrazolium (TNBT, black precipitate), 2- (4-iodophenyl) -5- (4-nitrophenyl) -3-phenyltetrazolium chloride (INT, red or orange precipitate) ), Tetrazolium blue (blue), nitrotetrazolium violet (violet) and 3- (
(5,5-dimethylthiazol-2-yl) -2,5-diphenyltetrazolium bromide (MTT, purple) and the like, but are not limited thereto. All tetrazolium substrates require glucose as a co-substrate. Glucose is oxidized and tetrazolium salts are reduced to form insoluble formazan to form a colored precipitate.

Β-galactosidase substrates include 5-bromo-4-chloro-3-indoyl-
Examples include, but are not limited to, β-D-galactopyranoside (X-gal, blue precipitate).

The precipitate associated with each of the above-listed substrates has a unique detectable spectral signature.

Antibodies that bind heavy metals can be used for reflection contrast, brightfield or darkfield imaging or immunostaining using electron microscopy. Such heavy metals include, but are not limited to, gold and silver (typically colloidal).

The following references, which are incorporated herein, provide further examples. J. M. Elias (1990) Immunohistopathology
y: A practical approach to diagnosis (
Practical method for diagnosis), ASCP Press (American Society)
ty of Clinical Pathologists), Chicago; F
. McGinty and F.M. E. FIG. Bloom (1983) Double immu
nostaining reviews distinctions amon
goopid peptidergic neurons in the
media basal hypothalamus (by double immunostaining,
Differences between opioid peptide-acting neurons in the medial basal hypothalamus were found), B
rain Res. 278: 145-153; Jowett (1997
) Tissue In situ Hybridization: Methods in Animal
Development, John Wiley & Sons, Inc., New York; J Histochem Cytochem December 1997, 45 (12
): 1629-1641.

Tissue stain for transmission microscope : Some of the tissue stains used for transmission light microscope are shown below: eosin, hematoxylin, orange G, light green SF,
Romanovsky-Giemsa, May-Grunwald, blue counterstain (Tre
vigen), ethyl green (CAS), foilgen naphthol yellow S
, Giemsa, methylene blue, methyl green, pyronin, naphthol yellow, neutral red, Papanicolaou stain (typically including a mixture of hematoxylin, eosin Y, light green SF, orange G and bismark brown), red counterstain stain B (Trevigen), Red counterstain C (Tr
evigen) and Sirius Red.

DNA ploidy stain for transmission microscope : The following describes D used for transmission light microscope.
Some of the NA ploidy stains are shown: Feulgen reagent (paralosaniline), galocyanine chrom alum, galocyanine chrom alum and naphthol yellow S, methyl green pyronin Y and thionin-Feulgen reagent.

Immunohistochemical stains for fluorescence microscope : Fluorescein, Rhodamine, Texas Red, Cy2, Cy3, Cy5, VECTOR Red, ELF ™ (enzyme-labeled fluorescein), Cy0, Cy0.5, Cy1, Cy1 .5, Cy3, C
y3.5, Cy5, Cy7, FluorX, calcein, calcein-AM, C
RYPTOFLUOR ™ (orange (42 kDa), tangerine (35
kDa), gold (31 kDa), red (42 kDa), Crimson (40
kDa)), BHMP, BHDMAP, Br-Oregon, Lucifer Yellow,
Alexa dyes, N- [6- (7-nitrobenz-2-oxa-1,3-diazol-4-yl) amino] caproyl] (NBD), BODIPY ™, boron dipyrromethene difluoride, Oregon green, MITOTRACKER (
TM) Red, DiOC ₇ (3), DiIC ₁₈ , phycoerythrin, phycobiliprotein BPE (240 kDa) RPE (240 kDa) CPC (264 k)
Da) APC (104 kDa), spectrum blue, spectrum aqua, spectrum green, spectrum gold, spectrum orange, spectrum red,
NADH, NADPH, FAD, infrared (IR) dye, cyclic GDP ribose (cG
DPR), chalcofluor white, tyrosine and tryptophan.

Tissue stain for fluorescence microscope : 4 ', 6-diamidino-2-phenylindole (
DAPI), eosin, fluorescein isothiocyanate (FITC), Hoechst 33258 and Hoechst 33342 (two bisbenzoimides), propidium iodide, quinacrine, fluorescein-phalloidin and resorufin.

DNA ploidy stain for fluorescence microscope : Chromomycin A3, DAPI, Acriflavine-Foilgen reaction, Auramine O-Foilgen reaction, ethidium bromide, propidium iodide, high affinity DNA fluorophores (POPO, BO
BO, YOYO, TOTO, etc.), green fluorescent protein (histone etc.) fused to DNA binding protein, ACMA, quinacrine orange and acridine orange.

Endogenous pigments : hemoglobin, myoglobin, porphyrin, hemosiderin and other ferrous pigments, lipofuscin, melanin, neuromelanin, celloids, fluorescent oxidation products of lipids / proteins, carotenoids, pyridines, flavin nucleotides.

The following are known to specifically bind relevant cytological markers and have been used as components in immunohistochemical stains currently used for research, and have limited However, some of the primary antibodies used for diagnosis of various diseases are summarized below. Anti-estrogen receptor antibody (breast cancer), anti-progesterone receptor antibody (breast cancer), anti-p53 antibody (multiple cancer), anti-Her-2 / neu antibody (multiple cancer), anti-EGFR antibody (epidermal growth factor, multiple cancer) ), Anti-cathepsin D antibody (cancer such as breast cancer), anti-Bcl-2 antibody (apoptosis cells), anti-E-cadherin antibody, anti-CA125 antibody (cancer such as ovarian cancer), anti-CA15-3 antibody (breast cancer),
Anti-CA19-9 antibody (colon cancer), anti-c-erbB-2 antibody, anti-P-glycoprotein antibody (MDR, multidrug resistance), anti-CEA antibody (carcinoembryonic antigen), anti-retinoblastoma protein (Rb) Antibodies, anti-ras oncoprotein (p21) antibodies, anti-Lewis X antibodies (also referred to as CD15), anti-Ki-67 antibodies (cell proliferation), anti-PCNA antibodies (multiple cancers), anti-CD3 antibodies (T cells), Anti-CD4 antibody (helper T cell), anti-CD5 antibody (T cell), anti-CD7 antibody (thymocyte, immature T cell, NK killer cell), anti-CD8 antibody (suppressor T cell), anti-CD9 / p24 antibody (ALL )
Anti-CD10 (also called CALLA) antibody (common acute lymphoblastic leukemia), anti-CD11c antibody (monocyte, granulocyte, AML), anti-CD13 antibody (myelomonocytic cell, AML), CD14 antibody (mature mononuclear cell, granulocyte), anti-CD15 antibody (Hodgkin's disease), anti-CD19 antibody (B cell), anti-CD20 antibody (B cell), anti-CD22 antibody (B cell), anti-CD23 antibody (activation B cells, CLL), anti-CD3
0 antibodies (activated T cells and activated B cells, Hodgkin's disease), anti-CD31 antibodies (angiogenic markers), anti-CD33 antibodies (myeloid cells, AML), anti-CD34 antibodies (endothelial stem cells, stromal tumors), CD35 antibody (dendritic cells), anti-CD38 antibody (plasma cells, activated T, B and myeloid cells), anti-CD41 antibody (platelets, megakaryocytes), anti-L
CA / CD45 antibody (leukocyte common antigen), anti-CD45RO antibody (helper, inducer T cell), anti-CD45RA antibody (B cell), anti-CD39, CD100
Antibody, anti-CD95 / Fas antibody (Apoptosis), anti-CD99 antibody (Ewing Salcoma marker, MIC2 gene product), anti-CD106 antibody (VCAM-
1: activated endothelial cells), anti-ubiquitin antibody (Alzheimer's disease), anti-CD71 (
Transferrin receptor) antibody, anti-c-myc (tumor protein and hapten) antibody, anti-cytokeratin (transferrin receptor) antibody, anti-vimentin (endothelial cell) antibody (B cells and T cells), anti-HPV protein (human papilloma virus) Antibody, anti-kappa light chain antibody (B cell), anti-lambda light chain antibody (B cell), anti-melanosome (HMB45) antibody (melanoma), anti-prostate specific antigen (PSA
) Antibodies (prostate cancer), anti-S-100 antibodies (melanoma, salveari, glial cells)
), Anti-tau antigen antibodies (Alzheimer's disease), anti-fibrin antibodies (epithelial cells), anti-keratin antibodies and anti-Tn antigen antibodies (colon cancer, adenocarcinoma and pancreatic cancer).

Table 2 below shows several marker sets with prognostic and / or therapeutic significance in various cancers:

[0202]

[Table 2]

Therefore, the breast cancer panel includes anti-ER, anti-PR, anti-Her2 / neu, anti-p5
3. Anti-Ki-67 and anti-CD31 immunohistochemical marker stains, DNA ploidy stains and H & E counterstains can be included.

The ovarian and / or endometrial cancer panels therefore have anti-Her2 / n
eu, anti-p53, anti-Ki-67 and anti-CD31 immunohistochemical stains, DNA ploidy stains and H & E counterstains can be included.

The prostate and / or bladder cancer panels therefore include anti-Ki-67, anti-p53, anti-CD31 and anti-retinoblastoma protein (Rb) marker stains, D
NA ploidy stains and H & E counterstains can be included.

Thus, a colorectal cancer panel can include anti-Ki-67, anti-p53, anti-CD31 and anti-p21 (ras tumor protein) marker stains, DNA ploidy stains and H & E counterstains.

The invention described herein has several advantages over conventional methods. That is,
High resolution imaging and very low and poor spectral resolution (typically 2
Two distinct spectral bands) can be combined to detect the two stains.
This means that a desired number of stains can be obtained using the spectral collection device described herein, characterized by high spatial and high spectral resolution, even when the stains used are very similar in spectrum. This is because they can be detected together.

Thus, in accordance with the present invention, clinicians use common histostains to determine multiple cytological markers for prognostic / therapeutic significance (eg, ER, PR, p53, her
-2 / neu, Ki-67 and CD31), can be detected simultaneously with significant changes in DNA ploidy and sample staining. Thus, clinicians can manage and monitor patient treatment with improved accuracy (eg, selection of appropriate treatment programs), reduce patient management costs through more accurate diagnosis, and achieve more efficient sampling. Hospital costs can be reduced through.

The scanning of the stained biological sample according to the present invention can be performed manually, as is well known in the art, for example, using a slide loading and scanning device.
It can be performed semi-automatically or automatically. For a complete DNA ploidy assay, typically 200 cells, preferably more, must be analyzed. Pattern recognition and spectral decomposition pattern recognition methods can be used to increase analysis and diagnostic accuracy.

The method of simultaneous in situ analysis of multiple immunohistochemical stains, tissue stains and DNA ploidy stains provided by the present invention using high spatial and high spectral resolution spectral imaging methods is described in the prior art. There are numerous advantages to First, multiple stains can be detected simultaneously, and are therefore efficient and economical. Second, while the prior art is not applicable to two or more stains for the same cell or tissue, multiple stains for the same cell or tissue can be detected simultaneously. Third, the prior art requires the use of very distinct dyes, while allowing the separation between very similar spectra.

EXAMPLES The reference, together with the above description, consists of the following examples, which illustrate the invention in a non-limiting manner.

Example 1 Measuring Apparatus and Its Performance FIG. 1 is incorporated by reference as fully listed herein.
On February 21, 1995, Cabib et al. No. 08 / 392,019 to U.S. Pat. No. 5,539,517, issued Jul. 23, 1996, which is incorporated herein by reference. It is a block diagram of a component.

This imaging spectrometer has a high spectral (Ca.4-14
nm) and spatial (Ca. 30 / M μm, where M is the effective microscope or pre-optical magnification) resolution, so that it is very well constructed to implement the method of the present invention.

Thus, the prior art imaging spectrometer of FIG. 1 generally comprises a collection optics indicated generally at 20, a one-dimensional scanner as indicated by block 22, a block 24.
An optical path difference (OPD) generator or interferometer, as indicated by, a one-dimensional or two-dimensional detector array as indicated by block, a signal processor and display as indicated by block.

A very essential component in the system 20 is the OPD generator or interferometer 24, which corresponds to a predetermined set of linear combinations of the spectral intensities of the light emitted from each pixel of the scene to be analyzed. Output modulated light. The output of the interferometer is concentrated on a detector array 26. Thus, to obtain all of the information needed to reconstruct the spectrum, all of the necessary optical phase differences are scanned simultaneously for all pixels in the viewing area. The spectra of all the pixels in the scene are therefore collected at the same time as the imaging information, thus allowing the analysis of the image in a real-time manner.

The device according to US Pat. No. 5,539,517 may be implemented in many different configurations.
In particular, the interferometer used may be combined with other mirrors as described in the related figures of US Pat. No. 5,539,517.

Therefore, other types of interferometers may be used in accordance with US Pat. No. 5,539,517. These include: (i) a moving interferometer in which the OPD changes to modulate the light, i.e., a Fabry-Perot interferometer with a scanning thickness; (ii) receiving light from the optical collection system and scanner; A Michelson-type interferometer that includes a beam splitter that splits the beam into two paths, (iii) within which an ODP, such as a four-mirror and beam splitter interferometer as further described in the referenced US patents. A Sagnac interferometer (see FIG. 14 of that specification) is optically combined with other optical methods that vary with the angle of incidence of the incoming radiation.

FIG. 2 illustrates an imaging spectrometer constructed according to US Pat. No. 5,539,517, which uses an interferometer such that the OPD varies with the angle of incidence of the incoming radiation. Light rays that enter the interferometer at a small angle to the optical axis undergo ODP that varies substantially linearly with this angle.

In the interferometer of FIG. 2, after being equilibrated by the optical sampling system 31, all of the radiation from the source 30 at all pixels is scanned by a mechanical scanner 32. The light then passes through a beam splitter 33 to a first reflector 34, then to a second reflector 35, passes through the beam splitter 33, and then reflects the light through a focusing lens 36 to an array of detectors 37 (eg, a CCD). You. This ray is interfered by the ray reflected by 33, then by the second reflector 35 and finally by the first reflector 34.

At the end of one scan, through all ODPs, all pixels are measured, so that the spectra of all pixels in the scene can be reconstructed by Fourier transform. Light rays parallel to the optical axis are corrected, and light rays at an angle (θ) to the optical axis undergo OPD which is a function of the thickness of the beam spreader 33, its reflection index, and the angle θ. OPD is proportional to θ for small angles. Calculate the spectrum of all pixels by making the appropriate transformations and by carefully recording.

In the configuration of FIG. 2, light rays incident on the beam splitter at an angle β (β = 45 ° in FIG. 2) travel through an interferometer with OPD = 0, while the general angle β−θ
The light ray incident on the light receiving element receives the OPD given by the equation (1).

(Equation 1) Where θ is the angular distance of the light beam from the optical axis or interferometer rotation axis with respect to the central site, t is the thickness of the beam spreader, and n is the refractive index of the beam spreader.

By scanning both the positive and negative angles with respect to the central part, we obtain a doubled interferogram for every pixel and remove the phase difference giving more accurate results in the Fourier transform calculation. It is understood from equation 1 that this helps. The width of the scan determines the maximum OPD reached, which is related to the spectral resolution of the measurement. The size of the angular step itself determines the OPD step defined by the shortest wavelength to which the system is sensitive. In fact, the sampling theorem [Chamberlain (1979) The princepi]
es of interferometric spectroscopy,
John Wiley and Sons, pp. 53-55], this OPD process must be smaller than half the shortest wavelength at which the system is sensitive.

Another parameter to take into account is the finite size of the detector elements in the matrix. Through focused light, this element delimits a finite OPD in the interferometer with the effect of involving an interferogram with a right angle relationship. This results in a decrease in the sensitivity of the system at short wavelengths, which falls to zero for wavelengths equal to or less than the OPD delimited by this factor. This implies that the modulation transfer function (MTF) condition is fulfilled, that is, the OPD ranged by the detector elements in the interferometer must be less than the shortest wavelength at which the device is sensitive. You have to make sure you don't.

Thus, an imaging spectrometer constructed in accordance with the invention disclosed in US Pat. No. 5,539,517 not only measures the intensity of light coming from every pixel in the field of view, The spectrum of each pixel in a predefined wavelength range is also measured. They also make better use of all the radiation emitted by each pixel in the field of view at any given time, and thus, as explained above, a clear reduction in frame time and / or spectrometer Allows a clear increase in the sensitivity of Such imaging spectrometers may include various types of interferometers and optical sampling and focusing systems, and thus are used for medical diagnostic and therapeutic and biological research applications, as well as for remote detection for geological and agricultural research, And may be used in various types of applications, including and the like.

As mentioned above, the imaging spectrometer according to the invention disclosed in US Pat. No. 5,539,517 uses Applied Spectral Imaging.
Ltd. , Industrial Park, Migdal Haemek,
Developed by Israel and denoted herein as SPECTRACUBE ^™ .

The SPECTRACUBE ^™ device has the following or better features, which are listed in Table 3 herein below.

[Table 3]

A SPECTRACUBE ^™ instrument optically coupled to a microscope is preferably used to analyze biological samples according to the methods of the present invention. However, any spectral imager, i.e., a filter (e.g., an auditory optical tunable filter (AOTF) or a liquid crystal tunable filter (LCTF)) and a dispersive element (e.g., grating or prism) based optical imager, or other Optical system data or multi-band light recovery devices (eg, Speicher®
. M. , Ballard S .; G. FIG. and Ward C.I. D. (1996)
Karyotyping human chromasomes by co
mbinatorial multi-flow FISH. Nature
Measured and stored in memory for subsequent correction and analysis of the spectrum of light illuminated by any point of interest located within its field of view, including devices according to the disclosure in Genetics, 12: 368-375). An instrument for storage can be used to obtain the required spectral data. Also, as described in US patent application Ser. No. 08 / 917,213, filed Aug. 25, 1997 and incorporated herein by reference as if fully set forth herein, a number of broadband ( (Which can be immobilized or modulated) can be used as a spectral data acquisition device according to the present specification. Accordingly, it is not intended to limit the scope of the present invention with respect to the use of particular types of spectral data acquisition instruments, but not particular types of spectral imagers.

SPECTRACU to separate and distinguish between very similar spectra
The power of the ^BETM device is displayed in FIGS. 3-5.

To illustrate this ability, first, chromosome 1 labeled with the fluorophores Texas-Red and Rhodamine, whose fluorescence spectra are very similar, respectively.
Note FIGS. 3A-C, including examples of interphase FISH measurements performed on chromosome 17-specific DNA. The chromosome 1 probe is a mid-satellite probe for the subtelomeric region of the chromosome, and was labeled with Texas-red which was linked to the DNA probe via biotin after hybridization. The chromosome 17 probe is an α satellite probe for the centromere region of the chromosome, and after hybridization, was labeled with rhodamine linked to a second DNA probe via digoxigenin. FIG. 3A is an image of the original, visible through a microscope. FIG. 3B shows the same sample after measurement and processing with a SPECTRACUBE ^™ instrument. On the other hand, FIG. 3C shows the fluorescence spectra of the Texas-Red (denoted T) and Rhodamine (denoted R) fluorophores.

As seen in FIG. 3C, the spectral peaks of Texas Red and Rhodamine were 1
They differ by only 5 nm, so it is very difficult to distinguish between them using a filter-based device.

As shown in FIG. 3A, looking at a color FISH image through a microscope,
Exact number of dots (marked 1-4) and probe types appearing in the image
Is not particularly high. On the other hand, as shown in FIG.
SPECTRACUBE with spectral advantages measured for pixels of ^TM The device is responsible for demonstrating the presence of dots, counting them accurately, and
And distinguish small sets with a high level of confidence due to small spectral differences between them
It is possible to As shown in FIG. 3C, Texas-Red and Rhodami
The localization of probe-specific fluorescence is determined with high accuracy by artificial coloring of the fluorescence.
Where dots 1 and 2 are of Texas-Red, and dots 3 and
4 is for rhodamine.

FIGS. 4A-B show hybrid forms of interfission DNA with six different probes.
An example of post-completion FISH measurement is provided. FIG. 4A shows the original as seen through a microscope.
Represents the image of an object. FIG. 4B shows the SPECTRACUBE of all detected sets. ^TM Represents the same image following instrument measurement, spectral processing and artificial coloring
. FIG. 4C is SPECTRACUBE.^TMThrough a triple dichroic filter
Represents the spectra of the six fluorophores after hybridization as detected by
1, 8, 10, 11, 17, and X according to the labeled chromosomes, respectively.
Noted). For more information on fluorophores, probes and chromosomes, see
The following description, Table 3 et seq. And Chroma Corp. Catalog number 61502
checking ...

The difficulty in distinguishing from each other by eye or even using a simple RGB color measurement is evident from FIG. 4A, which shows the original RGB image of the interphase nucleus.

The experimental observer can best detect three of the six different colors. However, FIG. 4B represents the same sample shown in FIG. 4A after processing the spectral data with the classification algorithm, highlighting the resulting dots with artificial colors, orange, cyan, blue, yellow, green and red. On the other hand, the background was black, giving an artificial color. As observed, it is possible to look at all six sets of fluorophores and easily distinguish the differences between the sets.

It is further noted that one set, one highlighted in blue, is barely noticeable by eye or by use of a color camera, but is detected after subjecting to a background subtraction algorithm on the spectral cube. Should (compare FIGS. 4A and 4B)

The probes used were five alpha satellite probes for the centromere of chromosomes 8, 10, 11, 17 and X, and mid satellite probes for the subtelomeric region of chromosome 1. The fluorophore and DAPI counterstain (background) used to label each of the above chromosomes, their emission peaks and artificially labeled color classifications are summarized in Table 4 below.

From the normalized spectral characteristics of each of the six fluorophores shown in FIG. 4C, due to the wide overlap between their spectra, a filter-based device that measures over a few relatively wide spectral ranges has been developed. It becomes clear that it is certainly not possible to distinguish between different probe species. Such a device relies more on absorption measurements of the intensity of each probe and is therefore more affected by background signals and noise. It should be further noted that the spectral overlap also occurs with the auto-fluorescence emanating from the cells themselves. Again, the validity of the spectral information for each pixel allows the elimination of the autofluorescent contribution, yielding more accurate results.

[Table 4] ¹ Vysis, Downers Grove, IL, U.S.A. S. As a labeled deoxynucleotide. Conjugated to a prehybridized digoxigenin containing ^two probes via an anti-digoxigenin antibody. ³ Fluorescein-
5-iso-thiothionate, conjugated to prehybridized biotin containing probe via anti-biotin antibody. ⁴ ', 6-diamidino-2-phenylindole used for ⁴ counterstaining.

FIGS. 5A-C are disclosed in US patent application Ser. No. 09 / 025,131, filed Feb. 17, 1998, which is hereby incorporated by reference as if fully set forth herein. 4 shows the results of hybridization of normal male chromosomes using a combinatorial hybridization approach in accordance with the outline provided.

FIG. 5A is an RGB image of the chromosome width obtained using the RGB algorithm. The RGB algorithm integrates the optical signals over the spectral width of the CCD array (eg, 400 nm to 760 nm) to obtain an RGB image of the chromosome, where each pixel is represented by a red (R), green (G), and blue ( Three specific gravity functions corresponding to the tristimulus response function to B), ｛w _r (λ
), W _g (λ), w _b (λ)} according to the combination of red, green and blue intensities.

In FIG. 5A, the specific gravity function w_r, W_g, W_bIs a simple square gravity function
Yes, w_r(Red) λ₁= 640 nm and λ₂= 750 nm, w_g(Green) λ ₁ = 555 nm and λ₂= 640 nm, w_b(Blue) λ₁= 450 nm and
λ₂= 555 nm. Simple weighting function w_r, W_g, W_bIntegrate the chromo
An RGB image of the soma was generated.

FIG. 5B is a chromosome width classification image of FIG. 5A. FIG. 5C is a karyotype derived from the chromosome width of FIG. 5B. The classification image is calculated by a classification algorithm in which each pixel is classified according to its spectrum. One of the most important analysis algorithms is a spectrum-based classification algorithm that can identify and enhance a large number of different spectra in an image with classified colors. This allows for the assignment of a specific classified color to all human chromosomes based on their spectra. This algorithm assumes that a control spectrum of each chromosome has been measured and stored in a computerized control library. A distinguishing color is assigned to each pixel in the image, as defined above, for example by a least squares error algorithm, according to the color assigned to the control spectrum that is most similar to the spectrum at a given pixel.

Example 2 Analysis and Display of Results General: A spectral image is a three-dimensional array of data, I (x, y, λ), that links spectral information with the spatial organization of the image. As such, due to its dimensions, a spectral image is a data set called a spectral cube, which is difficult to obtain in other ways, and makes extraction and quantitative evaluation of certain features even impossible in some cases. enable. Because both spectrometers and digital image analysis are well-known areas covered by a large amount of literature [eg, Jain
(1989) Fundamentals of Digital Image
Processing, Prentice-Hall International
nal], the following discussion will focus primarily on the benefits of concatenating spectral and imaging information in a single data set, ie, a spectral cube.

One possible type of analysis of a spectral cube is to use spectral data and spatial data separately, ie, apply spectral algorithms to spectral data, and apply two-dimensional image processing algorithms to spatial data. It is to apply to.

As an example of a spectral algorithm, the homology between the control spectrum and the spectra of all pixels is computed (ie, homology mapping), so that the intensity at each pixel is equal to the degree of “homology”. Consider an algorithm that results in a proportionally gray (or other color) scale image (ie, a homology map). This grayscale image can then be further analyzed using image processing and computer imagination techniques (eg, image enhancement, pattern recognition, etc.) to extract desirable features and parameters. In other words, homology mapping computes the integral of the absolute value of the difference between the spectra of each pixel of the spectral image with respect to the reference spectrum (already stored in the library or of the same or a different spectral image). Pixels), and displaying gray-level or pseudo-color (black and white or color) images, where light pixels correspond to small spectral differences and dark pixels correspond to large spectral differences. And vice versa.

Similarly, the classification mapping was performed with the same calculations as described for the homology mapping, and several spectra were taken as control spectra and their separation was determined to be most similar to one of several control spectra. , Each pixel of the displayed image was dyed with a different predetermined pseudo color.

A spectral image algorithm based on non-separable operations, ie, an algorithm that includes both local spectral information and the spatial relationship between neighboring pixels (one of these algorithms is principal component analysis, as shown below) It is also possible to use it.

Working with arbitrary three-dimensional (3D) data is referred to as a spectral cube (ie, I (
One of the basic needs that commonly arise when structuring (x, y, λ)) is to visualize structuring the data in a meaningful way. In general, the morphological data, D (x, y, z), obtained by a confocal microscope, where each point represents the intensity at a different position (x, y, z) in three-dimensional space. Unlike such other types of 3D, a spectral image is an array of images representing the intensity of the same two-dimensional plane (ie, sample) at different wavelengths. From this, the two most intuitive methods for viewing a spectral cube of data are the single pixel or set of pixels as a function of wavelength in the image plane (spatial data) or three-dimensional peak-valley representation. Is to look at any of the intensity. In general, the image plane is used to display either intensity measured at any single wavelength or a grayscale image that is the result of applying a spectral analysis algorithm on the desired spectral region at every image pixel it can. In general, the spectral axis can be used to represent the resulting spectrum of some spatial operations (eg, averaging the spectrum) performed near the desired pixel.

For example, a spectral image was used as a grayscale image similar to the image that could be obtained from a simple monochrome camera, or using one or several artificial colors to enhance and map key features. It can be displayed as a multicolor image. Since such cameras simply integrate the optical signal over the spectral width of the CCD array (eg, 400 nm to 760 nm), an “equal amount” monochrome CCD camera image should be integrated along the spectral axis as follows: Can be calculated from a 3D spectral image database.

(Equation 2)

In Equation 2, w (λ) is a general weighted response function that provides maximum flexibility in calculating various grayscale images, all of which are appropriately weighted over some spectral width. Based on the integrated spectral image.
For example, three different density function corresponding to the tristimulus response functions for red Equation 2, respectively (R), green (G) and blue _{(B), {w r (} λ), w g (λ),
When evaluated with w _b (λ)}, a conventional RGB color image can be displayed.
It is also possible to display meaningful non-traditional (pseudo) color images.
FIG. 6 illustrates an example of the capabilities of this simple algorithm. Of the target spectrum
Spectrum color image corresponding to the specific gravity function - to be a Gaussian function to distribute inward _{"{w r, w g, w} b} to consider choosing the pseudo obtained are displayed in this case It is possible to emphasize only the data in the regions and more clearly detect the spectral differences in these three regions.

Point Operations: Point operations are defined as being performed on a single pixel (ie, not including one or more pixels in time). For example, in a grayscale image, point manipulation may map the function of each pixel (intensity function) into other intensities according to a previously determined transformation function. A particular case of this type of transformation is the multiplication of the intensity of each pixel by a constant. Further examples include homology and taxonomy mapping as described herein above.

The concept of point manipulation can also be extended to spatial images. Here, each pixel has its own intensity function (spectrum), that is, an n-dimensional vector V ₁ (λ); λ∈ [λ ₁ , λ _n ]. The point operations applied to the spatial image are transformed functions

(Equation 3) Scalar (ie, intensity value) the spectrum of each pixel according to
Can be defined as mapping to

Assembling a grayscale image according to Equation 3 is an example of this type of point operation. In a more general case, the point operation converts the spectrum (vector) of each pixel to a transformation function

(Equation 4) According to N < n in the expression, the image is mapped to another vector.

In this case, the spectral image is converted to another spectral image.

Here, the definition of point operations can be extended to include operations between corresponding pixels of different spectral images. An important example of this type of algorithm is optical density analysis. Optical density is used to emphasize and graphically represent the area of interest that is being studied spectrally in a higher dynamic range than the transmission spectrum. Optical density is related to transmission by logarithmic operation and is therefore always a positive function. The relationship between optical density and measured spectrum is Lambert-Beer law

(Equation 5) Where OD (λ) is the optical density as a function of wavelength, I (λ) is the measured spectrum, I ₀ (λ) is the measured control spectrum, and τ (λ) is the spectral transmission of the sample. is there. Equation 5 selects I ₀ (λ) from (1) the pixel in the same spectral cube for which the OD was measured, (2) the corresponding pixel in the second cube, and (3) the spectrum from the library. It is calculated for every pixel for every wavelength in the case.

It should be noted that the optical density does not depend on either the spectral response of the system being measured or the non-uniformity of the CCD detector. This algorithm is useful for mapping the relative concentrations, and, in some cases, the absorption concentrations of the absorbent in the sample, if their absorption and the thickness of the material are known.

Further examples include (i) adding, subtracting, multiplying a given spectrum,
An arithmetic function such as division and combinations thereof is applied to the spectrum of each pixel in the spectral image, where the resulting spectrum of each pixel is the sum, difference, between each spectrum of the first cube and the selected spectrum. Various straight lines, such as producing a new spectral cube, product ratio or combination, (ii) applying a given scalar to the spectrum of each pixel of the spectral image by an arithmetic function as described above. Includes, but is not limited to, combinatorial analysis.

[0257] Such a straight line combination is used for background subtraction, for example, in which the spectrum of a pixel located in the background area is subtracted from the spectrum of each pixel, and when the spectrum measured before analyzing the sample is a spectrum. It may be used for calibration procedures such as those used to split the spectrum of each pixel in the image.

Other examples include ratio image calculations and display as gray level images.
This algorithm calculates the ratio between the intensities at two different wavelengths for every pixel of the spectral image and colors each pixel with a lighter or darker artificial color accordingly. For example, it colors the pixels lighter for higher ratios and darker for lower ratios (or vice versa), indicating a distribution of spectrally sensitive materials.

Spatial-spectral mixing operation: In all spectral image analysis methods described above, an algorithm is applied to spectral data. The importance of displaying spectrally processed data as images is almost qualitative and provides the user with a useful image. However, depending on the application, it is also possible to use useful imaging data in a more meaningful way by applying an algorithm that uses the spatial-spectral relationship inherent in the aerial image. Space-spectrum manipulation is the most powerful type of spectral image analysis algorithm. As an example, consider the following conditions:

The sample contains k-cell types stained by k-differential staining (the phrase “cells” herein is used both for biological cells and also as “areas within the field of view of the device”). use). Each stain has a different spectrum and binds to only one k-cell type. It is important to find the average intensity per cell for each one k-cell type. The following procedure can be used to perform this task. (1) classify each pixel in the image as belonging to one k + 1 class (k cell type plus background) according to its spectrum; (2) classify the image into various cell types and Counting the number of cells in the
(3) Sum the spectral energy contributed by each class and divide it by the total number of cells from the corresponding class.

This procedure makes use of both spectral and spatial data. Relevant spectral data takes the form of characteristic cell spectra (ie, spectral “characteristics”), while spatial data is data as various types of cells (ie, cell blots), many of which also appear to the eye. Consists of In the above situation, cells can be distinguished by characteristic spectral characteristics. Therefore, the preferred point manipulation may be performed to produce a composite image where each pixel is assigned one of the k + 1 values. The spectra of the different cell types are s _i (λ); i = 1,
, K, λ∈ [λ ₁ , λ _n ], and the measured spectrum at each pixel (x, y) is s _{x, v} (λ), λ∈ [λ ₁ , Λ _n ]
Then, the following algorithm is a possible method of classification (step 1 above).

Let e ² _{i be} a deviation of the measured spectrum from the known spectrum of the staining attaching to cell type i. Then we adopted the least-squares definition of "distance",

(Equation 6) Where R _λ is the spectral region of interest. Each point [pixel (x, y)] in the image can now be categorized into one of k + 1 using the following definition.

(Equation 7)

Steps 2 and 3 above (calculation of image segmentation and average intensity) are simple using standard computer visual operations on composite images made according to the algorithms described herein in Equations 6 and 7.

Another approach is to measure the spectrum s _{x, y} (
λ) is a linear combination of k known fluorescence spectra, s _i (λ); i = 1, 2,.
, K. in this case,

(Equation 8) , C _k ], which is interpreted as a count vector C = [c ₁ , c ₂ ,..., C _k ]. i = 1,2, ......, interpretation of dF / _dc i = 0 for k (i.e. to find the value of _{c i} which minimize F) is determinant

(Equation 9) Where A is an element

(Equation 10) Is a square matrix of dimension k with

[Equation 11] Is a vector defined as

Arithmetic operations may be similarly applied to two or more spectral cubes and / or spectra of a given pixel or from a library. For example, applying an arithmetic operation between corresponding wavelengths of a corresponding set of pixels belonging to a first spectral cube of data and a second cube of data, e.g.
Consider obtaining a third cube that is the result of the data for purposes such as averaging two spectral cubes, tracking time change, and spectral normalization.

In many cases, the objects (eg, cells) present in a spectral image differ from one another in chemical composition and / or structure to some extent, especially when stained. A non-correlation analysis such as principal components analysis is used to emphasize these differences by producing a covariance or correlation matrix. An uncorrelated statistical analysis is indicated by extracting uncorrelated data over a larger amount of data and by means in their correlated portion. There are many related statistical correlation methods. Examples include principal component analysis (PC
A), including, but not limited to, normative variable analysis and specific value decomposition, the PCA of these methods is probably more common and, as defined above, the decorrelation of the spectral data Are used according to the invention. However, all decorrelation statistical methods, including those listed above, are related to each other,
It is possible that the fact that there is no intention to limit the scope of the invention to the use of any particular decorrelation method. In particular, it is not intended that the scope of the invention be limited to the use of principal component analysis, as other uncorrelated statistical methods may be used. Information regarding the use and operation of the uncorrelated statistical methods listed above can be found in R.A., both of which are fully incorporated herein by reference. A. Johnson and D.S. W. Witch
en, "Applied Multivariance Statistical
al Analysis, third edition, Prentice
Hall (1992) and T.W. W. Anderson, An Intro
Duction to Multivariance Statistical
Analysis, second edition, Wiley and
Sons (1984).

Further, as will become apparent from the description below, the means of decorrelation statistical methods may be performed with various refinements. Since the inventive concept does not depend on any particular refinement, the scope of the present invention is not intended to be limited to any particular refinement as described below.

A brief description of principal component analysis using covariance is provided below. For further details on principal component analysis, see Martins and Naes (1989).
) Multivariate Calibration, John Wile
y & Sons, Great Britain; and Esbensen et al., Eds, (1994) Multi variance analysis-
In fact, computer-assisted modeling as CAMO and Unscramble
See ler's User's Guide, Trondheim, Norway.

Thus, the intensity of a pixel of the image at wavelength λ _i (i = 1,..., N) is now considered as a vector whose wavelength is equal to the number q of pixels. Since there is one, or N total, of these vectors for every wavelength of measurement, these vectors can be arranged in a matrix B 'with q columns and N stages.

(Equation 12)

The mean for each stage of the matrix B 'is defined as follows:

(Equation 13) And the second standardization matrix B is

[Equation 14] Is defined as

A covariance matrix C is a matrix of dimension N × N B: C = B^T-Define as B. C is diagonal
And CV_i= Μ_iV_i(Where V_iAnd N are orthogonal unit vectors, μ _i Is the i-th unit vector V_iIs the eigenvalue representing the variable in the direction of
Associated eigenvectors and eigenvalues. In general, the least components are
Represents the highest variable as a function of the

The product BV _i (i = 1,..., N) is the projection of the spectral image on the orthogonal principal components. These are vectors with q elements (q = number of pixels),
Can be displayed separately as black and white images. These images may exhibit features that are not apparent from the regular black and white images filtered at a particular wavelength or wavelength range.

Example 3 Materials and Methods Breast Cancer Samples : All samples were obtained from 79 women with moderately differentiated invasive ductal carcinoma. Samples were transected and prepared for staining according to conventional protocols.

Staining protocol : Staining was performed according to standard Vantana or DAKO automated immunostaining protocols.

Measurement : Microscope (Nikon Ec) connected to SPECTRACUBE ^™ instrument
(Lipse E-800) was adjusted to the maximum voltage (12 V) for Koehler illumination with a transmission light lamp for the most stable illumination. Intermediate density and color filters (FG3 and anti-reflection filters) were introduced in the light path to adjust the intensity and spectral color balance. The SPECTRACUBE acquisition parameters were 300 frames, 512 virtual frames, wavelength width 440 to 760 nm, 176 ms / frame.

First, the “pure dye”, ie, hematoxylin, DAB, AEC, and Fast Red (each only a colorant) spectral cube is captured and a representative spectrum therefrom is used to form a pure dye spectral library. did.

Next, a spectrum cube of each sample was captured. And SPECTR
ACUBE ^TM instruction manual in SPECTRACUBE ^TM algorithm as described in the SpyView, was used to obtain each spectral component, threshold two yuan of image and RGB image thereof mixed classification image, the gray level image.

Cervical cancer samples : Middle-aged female pap smears were collected by conventional procedures.

Staining Protocol : Staining is described in G., et al., Essentially incorporated by reference as fully listed herein. Papanicolaou (1942) A
new procedure for staking vaginals
means. Science 95: 438-439.

Measurement : Microscope (Nikon Ec) connected to SPECTRACUBE ^™ instrument
(Lipse E-800) was adjusted to the maximum voltage (12 V) for Koehler illumination with a transmission light lamp for the most stable illumination. Intermediate density and color filters (FG3 and anti-reflection filters) were introduced in the light path to adjust the intensity and spectral color balance. The SPECTRACUBE acquisition parameters were 300 frames, 512 virtual frames, wavelength width 440 to 760 nm, 176 ms / frame.

First, “pure dyes”, namely, harris hematoxylin, eosin, orange G
And Light Green SF, Bismarck Brown Y (each colorant only) spectral cubes were captured and representative spectra from them were used to form a pure dye spectral library.

[0282] A spectral cube of the cervical cancer sample was then captured. And SPECTRA
The SPECTRACUBE ^™ algorithm, SpyView, as described in the CUBE ^™ instruction manual, was used to obtain RGB images, gray level images of the respective spectral components, threshold binarized images and their mixed classification images.

Example 4 Experimental Results Two well-characterized experimental models, breast cancer and cervical cancer samples, have helped to demonstrate the feasibility and practicality of the method according to the present method.

[0284] Figure 7, six single colorant stain breast cancer samples from SPECTRACUBE ^TM device two histological staining was measured using (hematoxylin and eosin) and four immunohistochemical staining (DAB, Fast Red, AEC and BCIP /
NBT). The peak wavelength is shown on the right. Note that each stain has a characteristic spectrum, which allows for simultaneous co-detection of the spectral components associated with each of them, as further exemplified herein below.

FIGS. 8-11 represent such an experiment. All images were measured using the SPECTRACUBE ^™ instrument and its various measurement and analysis algorithms as described herein above.

FIGS. 8A-E represent images of breast cancer that have already been independently determined to be ER (+) / PR (+). Samples were co-stained with histologically stained hematoxylin and immunohistochemical staining, anti-ER-DAB. FIG. 8A represents an RGB image from a sample using the RGB algorithm described herein above in connection with FIG. FIG.
BD show binary images of hematoxylin, DAB and AEC spectral components, respectively. These binary images were obtained by threshold grayscale images showing the intensity of each component at each pixel. FIG. 8D
Serves as a mimic for PR (-) tumors. FIG. 8E shows a classification overlay image, where the spectral components are highlighted in red, green and blue, respectively. It should be noted that, as expected, the AEC spectral components are not detectable, and the hematoxylin and DAB spectral components are easily detectable. The classification image assists the pathologist in assessing the presence / absence / aggression level / diagnosis and / or diagnosis of the investigated cancer cells or tissues.

FIGS. 9A-E show images of breast cancer cells that have already been independently determined to be ER (+) / PR (+). This sample was also co-stained with histological staining hematoxylin and immunohistochemical staining, anti-PR-AEC. FIG. 9A shows an RGB image of the sample. 9B-D represent binarized images of hematoxylin, DAB and AEC spectral components, respectively, while FIG. 9E represents a classification overlay image;
Thus, the spectral components are enhanced in red, green, and blue, respectively. As expected, no DAB spectral components were detectable and hematoxylin and AE
It should be noted that both C spectral components are easily detectable.

FIGS. 10A-E show that the ER (+) / PR (+) are already independent, as in the previous sample.
1 represents an image of a breast cancer cell determined to be. However, in this case the samples were co-stained with immunohistochemical staining, anti-PR-fast red, replacing the histological staining hematoxylin and AEC already used. FIG. 10A shows the RGB values of the sample.
Represents an image. FIGS. 10B-D show binary images of hematoxylin, DAB and Fast Red spectral components, respectively. FIG. 10E shows a classification overlay image, where the spectral components are highlighted in red, green, and blue, respectively. Note that, as expected, DAB spectral components are not detectable in this sample, and that both hematoxylin and fast red spectral components are easily detectable, except for artificial crystals, especially formed in the lower part of the region. Should.

FIGS. 11A-E show images of breast cancer cells that have already been independently determined to be ER (+) / PR (+). This sample was also co-stained with histologically stained hematoxylin and immunohistochemical staining, anti-ER-DAB, anti-PR-fast red. FIG. 11A shows an RGB image of the sample. FIGS. 11B-D show binarized images of hematoxylin, DAB and Fast Red spectral components, respectively. FIG.
E represents the classification overlay image, where each of the spectral components is
Highlighted in red, green and blue. Areas co-stained with anti-ER-DAB and anti-PR-fast red are shown in yellow. It should be noted that, as expected, hematoxylin, DAB and Fast Red spectral components are easily detectable.

FIGS. 12A-F show milk independently determined to be ER (+) / PR (+).
1 shows an image of a cancer cell. This sample contains histologically stained hematoxylin and eosin.
And immunohistochemical staining, anti-ER-DAB, anti-PR-fast red
Was. FIG. 12A shows an RGB image of the sample. FIGS. 12B-E show hematoki, respectively.
Binary image of cylin, eosin, DAB and fast red spectral components
Represents FIG. 12F shows a classification overlay image where the spectrum
Ingredients are highlighted in blue, purple, green and red, respectively, SPECTRACUBE ^TM Figure 4 illustrates the ability of the device to separate four different spectral components. Accordingly
In this example, the nucleus (hematoxylin), cytoplasm (eosin), est
Rogen receptor (ER, detected by anti-ER / HRP / DAB) and progester
Showing staining for non-receptors (PR, detected with anti-PR / HRP / DAB)
2 shows a section of a breast cancer tissue obtained. In the image, some ER (+) cells
Fewer PR (+) cells were shown. In different areas of the entire slide,
ER (+) / PR (+)) and interductal (ER (+) / PR (-)) cancer
. This image shows the vessels and therefore mainly ER (+) / PR (-) tumor cells.
A vesicle was expected.

FIG. 13 shows SPECTRACU from a cervical cancer sample stained with five single colorants.
FIG. 4 represents the non-standardized spectrum of five histological stains (Harris Hematoxylin, Eosin, Orange G, Light Green SF, Bismarck Brown Y) measured using a ^BETM instrument. The peak wavelength is shown on the right.

FIGS. 14A-G show histological stains, Harris Hematoxylin, Eosin, Orange G, Light Green SF, Bismarck Brown Y, collectively forming what is known in the art as Papanicolaou (Pap) stain. 1 represents an image of a co-stained cervical cancer sample. FIG. 14A shows an RGB image of the sample. 14B-F show binary images of Harris Hematoxylin, Eosin, Orange G, Light Green SF and Bismarck Brown Y spectral components, respectively. FIG. 14E shows a classification overlay image where the spectral components are blue,
It is highlighted in pink, orange, green, and gray and forms a versatile classified overlay image indicated by their combination. It should be noted that each stain used had a unique staining pattern that could only be analyzed due to the high spectral and spatial resolution of the SPECTRACUBE ^™ instrument used.

The data presented herein demonstrate the utility of devices having high spectral and spatial resolution in analyzing biological samples co-stained with multiple colorants.

The algorithms and representations used under this example can be replaced by a number of other algorithms and representations, one example of which is incorporated by reference as if fully set forth herein. Those skilled in the art will appreciate that the linear decomposition algorithm is described in U.S. Patent No. 08 / 984,990. Using this algorithm, the analyst would be able to obtain another very similar direct result.

Although the present invention has been described in relation to particular embodiments thereof, it is evident that many modifications, changes and variations will be appreciated by those skilled in the art. It is therefore intended that all such modifications, alterations, and variations that fall within the spirit and scope of the appended claims be included.

[Brief description of the drawings]

FIG. 1 is a block diagram showing the main components of an imaging spectrometer configured according to US patent application Ser. No. 08 / 392,019 (prior art).

FIG. 2 shows a non-moving interferometer, ie, a Sagnac interferometer, used in an imaging spectrometer according to US patent application Ser. No. 08 / 392,019 (prior art).

FIG. 3A shows a mesophase FISH performed with two different probes conjugated to Texas Red and Rhodamine.

FIG. 3B shows mesophase FISH performed using two different probes conjugated to Texas Red and Rhodamine.

FIG. 3C shows mesophase FISH performed with two different probes conjugated to Texas Red and Rhodamine.

FIG. 4A. SPECTRAC for 6 different probes, each labeled with a different fluorophore.
Interphase FISH performed using UBE ™ equipment.

FIG. 4B. SPECTRAC for 6 different probes, each labeled with a different fluorophore.
Interphase FISH performed using UBE ™ equipment.

FIG. 4C. SPECTRAC for 6 different probes, each labeled with a different fluorophore.
Interphase FISH performed using UBE ™ equipment.

FIG. 5A shows the results of hybridization between 24 chromosome paints and a normal male chromosome spread obtained using a SPECTRACUBE ™ apparatus.

FIG. 5B shows the results of hybridization between 24 chromosome paints and a normal male chromosome spread obtained using a SPECTRACUBE ™ apparatus.

FIG. 5C shows the results of hybridization between 24 chromosome paints and a normal male chromosome spread obtained using a SPECTRACUBE ™ apparatus.

FIG. 6 illustrates the definition of pseudo RGB (red, green and blue) colors to enhance selected spectral ranges.

FIG. 7. SPECTRACUBE ™ from breast cancer samples stained with six single stains
Two tissue stains (hematoxylin and eosin) and four immunohistochemical stains (DAB, Fast Red, AEC and BCIP / N) were measured using the instrument.
3 shows an unnormalized spectrum of (BT).

FIG. 8 is a diagram showing an image of a breast cancer sample known to be ER (+) / PR (+) previously co-stained with a tissue stain, hematoxylin, and an immunohistochemical stain, anti-ER-DAB.

FIG. 9 is a diagram showing an image of a breast cancer sample which has been known to be ER (+) / PR (+) in advance, which has been co-stained with a tissue staining agent hematoxylin and an immunohistochemical staining agent anti-PR-AEC.

FIG. 10 is a diagram showing an image of a breast cancer sample previously known to be ER (+) / PR (+) co-stained with a tissue stain, hematoxylin, and an immunohistochemical stain, anti-PR Fast Red.

FIG. 11. Tissue stain hematoxylin and immunohistochemical stain anti-ER-DAB and anti-PR
FIG. 9 shows images of breast cancer samples previously known to be ER (+) / PR (+) co-stained with Fast Red.

FIG. 12. Hematoxylin and eosin histostains and anti-ER-D immunohistochemical stains
FIG. 4 shows images of breast cancer samples previously known to be ER (+) / PR (+) co-stained with AB and anti-PR-Fast Red.

FIG. 13 shows five tissue stains (Harris hematoxylin, eosin, orange G, light green SF and light green) measured using a SPECTRACUBE ™ instrument from cervical cancer samples stained with five single stains. It is an unnormalized spectrum of Bismarck Brown Y).

FIG. 14: Harris hematoxylin, eosin, orange G, a tissue stain that together form what is known in the art as a Papanicolaou stain
1 shows images of cervical cancer samples co-stained with Light Green SF and Bismarck Brown Y.

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G01N 21/78 G01N 21/78 C 33/533 33/533 33/58 33/58 A (81) Designated country EP (AT, BE, CH, CY, DE, DK, ES, FI, FR, GB, GR, IE, IT, LU, MC, NL, PT, SE), OA (BF, BJ, CF, CG, CI, CM, GA, GN, GW, ML, MR, NE, SN, TD, TG), AP (GH, GM, KE, LS, MW, SD, SL, SZ, UG, ZW), EA ( AM, AZ, BY, KG, KZ, MD, RU, TJ, TM), AE, AL, AM, AT, AU, AZ, BA, BB, BG, BR, BY, CA, CH, CN, CU, CZ , D , DK, EE, ES, FI, GB, GD, GE, GH, GM, HR, HU, ID, IL, IN, IS, JP, KE, KG, KP, KR, KZ, LC, LK, LR, LS, LT, LU, LV, MD, MG, MK, MN, MW, MX, NO, NZ, PL, PT, RO, RU, SD, SE, SG, SI, SK, SL, TJ, TM, TR TT, UA, UG, US, UZ, VN, YU, ZA, ZW (72) Inventors Kabib, Dario Israel, 23840 Timrat, Habrosh 7 (72) Inventors Buckwald, Robert Israel, 30095 Ramat Islay , Hadagan Street (without address) F-term (reference) 2G045 CB01 DA12 DA13 FA16 FB01 FB03 FB07 FB12 FB13 2G054 BB08 CA22 GB02 4B063 QA08 QA19 QQ02 QQ42 QQ79 QQ96 QR02 QR13 QR15 QR48 QS03 QS33 QS22 QS33

Claims (63)

[Claims] 1. A method for in-situ analysis of a biological sample, comprising: (a) a biological sample, wherein the first stain is a first immunohistochemical stain, a first tissue stain, and a first DNA ploidy stain. Wherein the second staining agent is selected from the group consisting of a second immunohistochemical staining agent, a second tissue staining agent, and a second DNA ploidy staining agent. A step of staining with a certain N kinds of staining agents, wherein N is an integer greater than 3, and (i) when the first staining agent is the first immunohistochemical staining agent, The second stain is the second tissue stain or the second DNA ploidy stain, and (ii) the second stain is used when the first stain is the first tissue stain. Is the second immunohistochemical stain or the second DNA ploidy stain, while When i) the first staining agent is the first DNA ploidy stain, the second staining agent is the second immunohistochemistry stain or the second staining agent,
And (b) collecting spectral data from the biological sample using a spectral data collecting device, wherein the spectral data collecting device and the N-type staining agent are:
Selecting such that spectral components associated with each of the N dyes can be collected. 2. The spectrum data acquisition device is selected from the group consisting of an interferometer-based spectrum data acquisition device, a filter-based spectrum data acquisition device, and a dispersive element-based spectrum data acquisition device. 2. The method according to 1. 3. The method of claim 1, wherein said first immunohistochemical stain and said second immunohistochemical stain each independently comprise a primary antibody. 4. The method according to claim 1, wherein the primary antibody is an anti-estrogen receptor antibody, an anti-progesterone receptor antibody, an anti-p53 antibody, an anti-Her-2 / neu antibody, an anti-EGF
R antibody, anti-cathepsin D antibody, anti-Bcl-2 antibody, anti-E-cadherin antibody, anti-CA
125 antibody, anti-CA15-3 antibody, anti-CA19-9 antibody, anti-c-erbB-2 antibody, anti-P-glycoprotein antibody, anti-CEA antibody, anti-retinoblastoma protein antibody,
Anti-ras oncoprotein antibody, anti-Lewis X antibody, anti-Ki-67 antibody, anti-PCNA
Antibody, anti-CD3 antibody, anti-CD4 antibody, anti-CD5 antibody, anti-CD7 antibody, anti-CD8 antibody, anti-CD9 / p24 antibody, anti-CD10 antibody, anti-CD11c antibody, anti-CD13 antibody, anti-CD14 antibody, anti-CD15 antibody, anti-CD19 Antibody, anti-CD20 antibody, anti-C
D22 antibody, anti-CD23 antibody, anti-CD30 antibody, anti-CD31 antibody, anti-CD33 antibody, anti-CD34 antibody, anti-CD35 antibody, anti-CD38 antibody, anti-CD41 antibody, anti-L
CA / CD45 antibody, anti-CD45RO antibody, anti-CD45RA antibody, anti-CD39 antibody, anti-CD100 antibody, anti-CD95 / Fas antibody, anti-CD99 antibody, anti-CD10
6, Anti-ubiquitin antibody, anti-CD71 antibody, anti-c-myc antibody, anti-cytokeratin antibody, anti-vimentin antibody, anti-HPV protein antibody, anti-kappa light chain antibody, anti-lambda light chain antibody, anti-melanosome antibody, anti-prostate specific Antibody, Anti-S-100
4. The method according to claim 3, wherein the method is selected from the group consisting of an antibody, an anti-tau antigen antibody, an anti-fibrin antibody, an anti-keratin antibody, and an anti-Tn antigen antibody. 5. The method of claim 3, wherein said primary antibody is labeled with a detectable component. 6. The method according to claim 5, wherein said detectable component is a fluorescent dye. 7. The fluorescent dye, wherein fluorescein, rhodamine, Texas red, Cy2, Cy3, Cy5, VECTOR red, ELF ™ (enzyme-labeled fluorescein), Cy0, Cy0.5, Cy1, Cy1.5, Cy3, Cy
3.5, Cy5, Cy7, FluorX, calcein, calcein-AM, CR
YPTOFLUOR ™ (orange (42 kDa), tangerine (35 kDa)
Da), gold (31 kDa), red (42 kDa), Crimson (40 kDa)
Da)), BHMP, BHDMAP, Br-Oregon, Lucifer Yellow, Alexa dyes, N- [6- (7-nitrobenz-2-oxa-1,3-diazol-4-yl) amino] caproyl] (NBD) , BODIPY ™, boron dipyrromethene difluoride, Oregon Green, MITOTRACKER ™ Red, DiOC ₇ (3), DiIC ₁₈ , phycoerythrin, phycobiliprotein BPE (240 kDa) RPE (240 kDa) CPC (264 kD)
a) APC (104 kDa), spectral blue, spectral aqua, spectral green, spectral gold, spectral orange, spectral red, N
ADH, NADPH, FAD, infrared (IR) dye, cyclic GDP ribose (cGD
7. The method of claim 6, wherein the method is selected from the group consisting of PR), Calcoflower White, Tyrosine and Tryptophan. 8. The method according to claim 5, wherein said detectable component is a non-fluorescent dye. 9. The method of claim 8, wherein said non-fluorescent dye comprises a heavy metal. 10. The method of claim 5, wherein the detectable component is an enzyme that catalyzes a colorimetric reaction of a substrate having a substantially insoluble color reaction product. 11. The method according to claim 10, wherein the enzyme is selected from the group consisting of alkaline phosphatase, horseradish peroxidase, β-galactosidase, and glucose oxidase. 12. The method according to claim 10, wherein said substrate is selected from the group consisting of an alkaline phosphatase substrate, a horseradish peroxidase substrate, a glucose oxidase substrate, and a β-galactosidase substrate. 13. The substantially detectable moiety is capable of emitting light or is substantially capable of being chemically reacted with a second reaction with at least one additional substrate and having a luminescent product. The method according to claim 5, which is an enzyme that catalyzes a luminescence reaction of a substrate having an insoluble reaction product. 14. The method of claim 13, wherein said enzyme is selected from the group consisting of alkaline phosphatase, horseradish peroxidase, β-galactosidase, and glucose oxidase. 15. The first substrate and the at least one additional substrate are each independently selected from the group consisting of an alkaline phosphatase substrate, a horseradish peroxidase substrate, a β-galactosidase substrate, and a glucose oxidase substrate. 14. The method of claim 13, wherein the method comprises: 16. The method of claim 1, wherein said first immunohistochemical stain and said second immunohistochemical stain each independently comprise a primary antibody and a signal amplification mechanism. 17. The signal amplification mechanism, wherein a secondary antibody capable of binding a constant region of the primary antibody, avidin or streptavidin capable of binding biotin conjugated to the primary antibody, and avidin conjugated to the primary antibody Or biotin capable of binding streptavidin.
7. The method according to 6. 18. The method of claim 14, wherein said secondary antibody, avidin, streptavidin and biotin are each independently labeled with a detectable component. 19. The method of claim 18, wherein said detectable component is a fluorescent dye. 20. The fluorescent dye, wherein fluorescein, rhodamine, Texas red, Cy2, Cy3, Cy5, VECTOR red, ELF ™ (enzyme-labeled fluorescein), Cy0, Cy0.5, Cy1, Cy1.5, Cy3, C
y3.5, Cy5, Cy7, FluorX, calcein, calcein-AM, C
RYPTOFLUOR ™ (orange (42 kDa), tangerine (35
kDa), gold (31 kDa), red (42 kDa), Crimson (40
kDa)), BHMP, BHDMAP, Br-Oregon, Lucifer Yellow,
Alexa dyes, N- [6- (7-nitrobenz-2-oxa-1,3-diazol-4-yl) amino] caproyl] (NBD), BODIPY ™, boron dipyrromethene difluoride, Oregon green, MITOTRACKER (
TM) Red, DiOC ₇ (3), DiIC ₁₈ , phycoerythrin, phycobiliprotein BPE (240 kDa) RPE (240 kDa) CPC (264 k)
Da) APC (104 kDa), spectrum blue, spectrum aqua, spectrum green, spectrum gold, spectrum orange, spectrum red,
NADH, NADPH, FAD, infrared (IR) dye, cyclic GDP ribose (cG
20. The method of claim 19, wherein the method is selected from the group consisting of DPR), chalcfluorowhite, tyrosine and tryptophan. 21. The method of claim 18, wherein said detectable component is a non-fluorescent dye. 22. The non-fluorescent dye comprises an alkaline phosphatase substrate, a horseradish peroxidase substrate, a glucose oxidase substrate,
22. The method of claim 21, wherein the method is selected from the group consisting of:-a galactosidase substrate. 23. The method of claim 18, wherein said detectable component is an enzyme that catalyzes a colorimetric reaction of a substrate having a substantially insoluble color reaction product. 24. The method of claim 23, wherein said enzyme is selected from the group consisting of alkaline phosphatase, horseradish peroxidase, β-galactosidase, and glucose oxidase. 25. The method of claim 23, wherein the substrate is selected from the group consisting of an alkaline phosphatase substrate, a horseradish peroxidase substrate, a β-galactosidase substrate, and a glucose oxidase substrate. 26. The detectable component catalyzes a luminescence reaction of a substrate having a substantially insoluble reaction product, which is capable of emitting light or directing a second reaction of a second substrate having a luminescence product. 19. The method according to claim 18, which is an enzyme. 27. The method according to claim 26, wherein said enzyme is selected from the group consisting of luciferase and aequorin. 28. The method of claim 26, wherein the first substrate and the second substrate are each independently selected from the group consisting of luciferin, ATP, coelenterazine, and Ca ^++. . 29. The first tissue stain and the second tissue stain are each independently 4 ′, 6-diamidino-2-phenylindole, eosin, fluorescein isothiocyanate, Hoechst 33258, Hoechst 33342, propidium ioda Id, quinacrine, fluorescein phalloidin, resorufin, hematoxylin, orange G, light green SF, Romanovsky-Giemsa, May-Grunwald, blue counterstain, ethyl green, foilgennaphthol yellow S, giemsa, methylene blue, methyl green, pyronin, The method according to claim 1, wherein the method is selected from the group consisting of naphthol yellow, neutral red, Papanicolaou stain, red counterstain C and Sirius red. 30. The first DNA ploidy stain and the second DNA ploidy stain each independently comprise chromomycin A3, DAPI, acriflavine-foylgen reaction, auramine O-foylgen reaction, ethidium bromide, propidium Iodide, high-affinity DNA fluorophore, green fluorescent protein fused to DNA binding protein, ACMA, quinacrine orange and acridine orange, Feulgen reagent, galocyanine chrom alum, galocyanine chrom alum and naphthol yellow S, methyl green pyronin Y and 2. The method of claim 1, wherein the method is selected from the group consisting of thionine- Feulgen reagents. 31. A method for in situ analysis of a biological sample, comprising: (a) a biological sample, wherein the first stain is a first immunohistochemical stain, a first tissue stain, and a first DNA ploidy stain. Wherein the second staining agent is selected from the group consisting of a second immunohistochemical staining agent, a second tissue staining agent, and a second DNA ploidy staining agent. A step of staining with a plurality of stains, provided that: (i) when the first stain is the first immunohistochemical stain, the second stain is the second tissue And (ii) when the first stain is the first tissue stain, the second stain is the second immunohistochemical stain. Or the second DNA ploidy stain, while (iii) the first stain is the second DNA ploidy stain. When a DNA ploidy stain, the second staining agent is the second immunohistochemistry stain or the second staining agent,
And (b) collecting spectral data from the biological sample using a spectral data collecting device, wherein the spectral data collecting device and the plurality of types of stains are respectively applied to the plurality of types of stains. Selecting the relevant spectral components to be collected. 32. The spectrum data acquisition device is selected from the group consisting of an interferometer-based spectrum data acquisition device, a filter-based spectrum data acquisition device, and a dispersive element-based spectrum data acquisition device. 31. The method according to 31. 33. The method of claim 31, wherein said first and second immunohistochemical stains each independently comprise a primary antibody. 34. The primary antibody is an anti-estrogen receptor antibody, an anti-progesterone receptor antibody, an anti-p53 antibody, an anti-Her-2 / neu antibody, an anti-EG
FR antibody, anti-cathepsin D antibody, anti-Bcl-2 antibody, anti-E-cadherin antibody, anti-C
A125 antibody, anti-CA15-3 antibody, anti-CA19-9 antibody, anti-c-erbB-2
Antibody, anti-P-glycoprotein antibody, anti-CEA antibody, anti-retinoblastoma protein antibody, anti-ras tumor protein antibody, anti-Lewis X antibody, anti-Ki-67 antibody, anti-PCN
A antibody, anti-CD3 antibody, anti-CD4 antibody, anti-CD5 antibody, anti-CD7 antibody, anti-CD8
Antibody, anti-CD9 / p24 antibody, anti-CD10 antibody, anti-CD11c antibody, anti-CD13
Antibody, anti-CD14 antibody, anti-CD15 antibody, anti-CD19 antibody, anti-CD20 antibody, anti-CD22 antibody, anti-CD23 antibody, anti-CD30 antibody, anti-CD31 antibody, anti-CD33
Antibody, anti-CD34 antibody, anti-CD35 antibody, anti-CD38 antibody, anti-CD41 antibody, anti-LCA / CD45 antibody, anti-CD45RO antibody, anti-CD45RA antibody, anti-CD39
Antibody, anti-CD100 antibody, anti-CD95 / Fas antibody, anti-CD99 antibody, anti-CD1
06 antibody, anti-ubiquitin antibody, anti-CD71 antibody, anti-c-myc antibody, anti-cytokeratin antibody, anti-vimentin antibody, anti-HPV protein antibody, anti-kappa light chain antibody,
Anti-lambda light chain antibody, anti-melanosome antibody, anti-prostate specific antigen antibody, anti-S-10
34. The method according to claim 33, wherein the method is selected from the group consisting of an antibody 0, an anti-tau antigen antibody, an anti-fibrin antibody, an anti-keratin antibody, and an anti-Tn antigen antibody. 35. The method of claim 33, wherein said primary antibody is labeled with a detectable component. 36. The method of claim 35, wherein said detectable component is a fluorescent dye. 37. The fluorescent dye may be fluorescein, rhodamine, Texas Red, Cy2, Cy3, Cy5, VECTOR Red, ELF ™ (enzyme-labeled fluorescein), Cy0, Cy0.5, Cy1, Cy1.5, Cy3, C
y3.5, Cy5, Cy7, FluorX, calcein, calcein-AM, C
RYPTOFLUOR ™ (orange (42 kDa), tangerine (35
kDa), gold (31 kDa), red (42 kDa), Crimson (40
kDa)), BHMP, BHDMAP, Br-Oregon, Lucifer Yellow,
Alexa dyes, N- [6- (7-nitrobenz-2-oxa-1,3-diazol-4-yl) amino] caproyl] (NBD), BODIPY ™, boron dipyrromethene difluoride, Oregon green, MITOTRACKER (
TM) Red, DiOC ₇ (3), DiIC ₁₈ , phycoerythrin, phycobiliprotein BPE (240 kDa) RPE (240 kDa) CPC (264 k)
Da) APC (104 kDa), spectrum blue, spectrum aqua, spectrum green, spectrum gold, spectrum orange, spectrum red,
NADH, NADPH, FAD, infrared (IR) dye, cyclic GDP ribose (cG
37. The method of claim 36, wherein the method is selected from the group consisting of DPR), Calcoflower White, Tyrosine and Tryptophan. 38. The method of claim 35, wherein said detectable component is a non-fluorescent dye. 39. The non-fluorescent dye comprises a heavy metal.
The method described in. 40. The method of claim 35, wherein the detectable component is an enzyme that catalyzes a colorimetric reaction of a substrate having a substantially insoluble color reaction product. 41. The method of claim 40, wherein said enzyme is selected from the group consisting of alkaline phosphatase, horseradish peroxidase, β-galactosidase, and glucose oxidase. 42. The method of claim 40, wherein said substrate is selected from the group consisting of an alkaline phosphatase substrate, a horseradish peroxidase substrate, a glucose oxidase substrate, and a β-galactosidase substrate. 43. The substantially detectable moiety is capable of emitting light or is substantially capable of being chemically reacted with a second reaction with at least one additional substrate and having a luminescent product. 36. The method according to claim 35, which is an enzyme that catalyzes a luminescence reaction of a substrate having an insoluble reaction product. 44. The method of claim 43, wherein said enzyme is selected from the group consisting of alkaline phosphatase, horseradish peroxidase, β-galactosidase, and glucose oxidase. 45. The first substrate and the at least one additional substrate are each independently selected from the group consisting of an alkaline phosphatase substrate, a horseradish peroxidase substrate, a β-galactosidase substrate, and a glucose oxidase substrate. 44. The method of claim 43, wherein 46. The method of claim 31, wherein said first immunohistochemical stain and said second immunohistochemical stain each independently comprise a primary antibody and a signal amplification mechanism. 47. The signal amplification mechanism, wherein a secondary antibody capable of binding the constant region of the primary antibody, avidin or streptavidin capable of binding biotin conjugated to the primary antibody, and avidin conjugated to the primary antibody Or biotin capable of binding streptavidin.
7. The method according to 6. 48. The method of claim 44, wherein said secondary antibody, avidin, streptavidin, and biotin are each independently labeled with a detectable component. 49. The method of claim 48, wherein said detectable component is a fluorescent dye. 50. The fluorescent dye, wherein fluorescein, rhodamine, Texas Red, Cy2, Cy3, Cy5, VECTOR Red, ELF ™ (enzyme-labeled fluorescein), Cy0, Cy0.5, Cy1, Cy1.5, Cy3, C
y3.5, Cy5, Cy7, FluorX, calcein, calcein-AM, C
RYPTOFLUOR ™ (orange (42 kDa), tangerine (35
kDa), gold (31 kDa), red (42 kDa), Crimson (40
kDa)), BHMP, BHDMAP, Br-Oregon, Lucifer Yellow,
Alexa dyes, N- [6- (7-nitrobenz-2-oxa-1,3-diazol-4-yl) amino] caproyl] (NBD), BODIPY ™, boron dipyrromethene difluoride, Oregon green, MITOTRACKER (
TM) Red, DiOC ₇ (3), DiIC ₁₈ , phycoerythrin, phycobiliprotein BPE (240 kDa) RPE (240 kDa) CPC (264 k)
Da) APC (104 kDa), spectrum blue, spectrum aqua, spectrum green, spectrum gold, spectrum orange, spectrum red,
NADH, NADPH, FAD, infrared (IR) dye, cyclic GDP ribose (cG
50. The method of claim 49, wherein the method is selected from the group consisting of: DPR), calcoflower white, tyrosine and tryptophan. 51. The method of claim 48, wherein said detectable component is a non-fluorescent dye. 52. The non-fluorescent dye comprises an alkaline phosphatase substrate, a horseradish peroxidase substrate, a glucose oxidase substrate,
52. The method of claim 51, wherein the method is selected from the group consisting of:-a galactosidase substrate. 53. The method of claim 48, wherein said detectable component is an enzyme that catalyzes a colorimetric reaction of a substrate having a substantially insoluble color reaction product. 54. The method of claim 53, wherein said enzyme is selected from the group consisting of alkaline phosphatase, horseradish peroxidase, β-galactosidase, and glucose oxidase. 55. The method of claim 53, wherein said substrate is selected from the group consisting of an alkaline phosphatase substrate, a horseradish peroxidase substrate, a β-galactosidase substrate, and a glucose oxidase substrate. 56. The detectable component catalyzes a luminescent reaction of a substrate having a substantially insoluble reaction product that is capable of emitting light or directing a second reaction of a second substrate having a luminescent product. 49. The method of claim 48, which is an enzyme. 57. The method of claim 56, wherein said enzyme is selected from the group consisting of luciferase and aequorin. 58. The method of claim 56, wherein said first substrate and said second substrate are each independently selected from the group consisting of luciferin, ATP, coelenterazine, and Ca ^++. . 59. The first tissue stain and the second tissue stain are each independently 4 ', 6-diamidino-2-phenylindole, eosin, fluorescein isothiocyanate, Hoechst 33258, Hoechst 33342, propidium ioda Id, quinacrine, fluorescein phalloidin, resorufin, hematoxylin, orange G, light green SF, Romanovsky-Giemsa, May-Grunwald, blue counterstain, ethyl green, foilgennaphthol yellow S, giemsa, methylene blue, methyl green, pyronin, 32. The method of claim 31, wherein the method is selected from the group consisting of Naphthol Yellow, Neutral Red, Papanicolaou Stain, Red Counterstain C and Sirius Red. 60. A method for in situ analysis of a biological sample, comprising: (a) staining a biological sample with at least four different immunohistochemical stains; and (b) using a spectral data collection device to obtain the biological sample. Collecting spectral data from a sample, wherein the spectral data collection device and the at least four immunohistochemical stains can be used to collect spectral components associated with each of the at least four immunohistochemical stains. And c. Selecting 61. A method for in situ analysis of a biological sample, comprising: (a) analyzing the biological sample, wherein at least one stain is an immunohistochemical stain and at least one further stain is a histostain or DNA ploidy. (B) collecting spectral data from the biological sample using a spectral data collecting device, wherein the spectral data collecting device and the at least three dyes are used. Selecting the species of stains such that spectral components associated with each of the at least three stains can be collected. 62. A method for in situ analysis of a biological sample, comprising: (a) a biological sample, wherein the first stain is an immunohistochemical stain, the second stain is a tissue stain, and the third stain is Staining with at least three types of stains, which is a DNA ploidy stain, and (b) collecting spectrum data from the biological sample using a spectrum data collection apparatus, wherein the spectrum data collection apparatus And selecting the at least three stains such that spectral components specifically associated with each of the at least three stains can be collected. 63. At least four different immunohistochemical stains, each immunohistochemical stain for staining a respective cytological marker,
An immunohistochemical composition wherein each immunohistochemical stain can be individually detected in the presence of all other immunohistochemical stains using a spectral data collection device.