BIOLOGICAL SPECIMEN EVALUATION METHODS USING CYTOLOGY AND IMMUNOLOGY

Abstract

Information on cytokines and cytology obtained from a biological specimen are combined as a method of predicting the risk that dysplasia will progress to cancer.

Claims

1. A method for estimating the risk of dysplasia or cancer progressing in a subject, the method comprising: (a) detecting and morphologically classifying dysplastic cells in a cytological preparation from a biological sample from the subject; (b) detecting the presence of inflammatory and anti-inflammatory cytokines in dysplastic or non-dysplastic cells; and (c) estimating the risk of dysplasia or cancer progressing in the subject based upon the ratio of the number of cells in the cytological preparation that are expressing at least one inflammatory cytokine to the number of cells that are expressing at least one anti-inflammatory cytokine, wherein the risk of progression decreases with an increasing value of this ratio.

2. The method of claim 1 wherein the anti-inflammatory cytokine comprises one or more of IL-10, TGF-, IL-4, IL-13 and interferon-alpha (INF-).

3. The method of claim 1 wherein the inflammatory cytokine comprises one or more of IL-12, IL-23, IL-1, tumor necrosis factor (TNF), interferon-gamma (IFN-) and granulocyte-macrophage colony stimulating factor (GMCSF).

4. The method of claim 1 wherein whether a cell is expressing a cytokine is determined qualitatively.

5. The method of claim 1 wherein whether a cell is expressing a cytokine is determined quantitatively.

6. The method of claim 1 wherein the expression of one or more anti-inflammatory cytokines by a morphologically dysplastic or cancerous cell is given greater weight in the estimation of the risk of progression than is the expression of the same anti-inflammatory cytokine(s) by a cell of the immune system.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawing(s) will be provided by the office upon request and payment of the necessary fee.

[0066] FIG. 1A to FIG. 1C shows results of IL-10 immunostaining on an IL-10 negative HSIL sample (D1H) with binding to antibodies Alexa-594 and Alexa-647; FIG. 1A is a scattergram of normal and dysplastic cells, (arrow points to red stained cells); FIG. 1B shows a cell stained with Alexa-594 (green) NSB; R-G correlation coefficient=0.088; FIG. 1C shows a cell stained with Alexa-647 (red) NSB; R-G correlation coefficient=0.67.

[0067] FIG. 2 is a scattergram showing results of IL-10 immunostaining on an IL-10 positive HSIL sample (D6H) with binding to antibodies Alexa-647 and Alexa-594; normal and dysplastic cells (arrow points to red stained cells).

[0068] FIG. 3A to FIG. 3H shows pseudocolor images of cells: FIG. 3A, FIG. 3B wherein non-specific binding was illustrated by color separated images of the same cell showing poor spatial correlation (<0.08) of the localization of the Alexa-594 (green), FIG. 3A and Alexa-647 (red) FIG. 3B labeled secondary antibodies. Blue denotes staining of the cell nucleus by POPRO; [FIG. 3C, FIG. 3D, FIG. 3E] show specific binding of FIG. 3C IL-10 positive HSIL clump; FIG. 3D HSIL clump594 stain; FIG. 3E HSIL clump647 stain; FIG. 3F, FIG. 3G, FIG. 3H shows specific binding of FIG. 3F IL-10 lymphocyte; FIG. 3G IL-10 lymphocyte594 stain; FIG. 3H IL-10 lymphocyte647 stain.

DETAILED DESCRIPTION

[0069] 1. Clinical Rationale

[0070] A method of using biological specimens to predict whether a patient is at an increased risk of developing cancer is disclosed.

[0071] In the absence of cancer, infection or cellular damage that has evoked a reparative response, the cells in a cytological specimen will have morphologies that are within normal limits, and indicators of an immune response such as the presence of T-cells, B-cells and macrophages will be very rare. In a similar manner, a cell undergoing normal repair of damage unrelated to cancer or an infection will show morphological changes characteristic of the repair process and some macrophages may be present, but T-cells and B-cells will be absent. However, if cancer, a significant precancerous condition, or an infection is present, the cell will exhibit distinctive morphological changes and evidence of an active immune response in the form of effector T-cells, Th1 helper T-cells, M1 macrophages and possibly other effector cells of the innate immune system. An aborted immune response, on the other hand is marked by dysplasia in combination with Treg and (possibly) Breg cells, Th2 helper T-cells and M2 macrophages as well as production of anti-inflammatory cytokines. Hence, the production of an anti-inflammatory cytokine, such as IL-10, by these cells is an indication of an aborted immune response and is an indicator that the local environment is in an immunosuppressed state that is permissive of progression of the adverse condition.

[0072] From a clinical perspective, the detection of cancer cells in a cytological specimen is considered to warrant immediate confirmatory follow up and aggressive intervention, while the response to dysplastic (pre-cancerous) cells in these specimens is more likely to be based upon the clinician's assessment of the risk that the dysplasia will progress to cancer. Standards of care worldwide largely dictate that high grade dysplasia (HSIL) represents a high risk of progression and therefore warrants immediate intervention, while lesser degrees of dysplasia represent lower risks and are responded to accordingly. This risk assessment is, however, complicated by the fact that a small fraction of ASCUS and LSIL are known to progress and that a significant portion of HSIL (60-90% by some estimates) spontaneously regress. A means of more accurately assessing in cytological specimens the risk of progression in a subject from where the specimen was obtained; is therefore desirable. The presence of cells of any type, secreting IL-10, is indicative that the patient is at high risk for progression, and is independent of the degree of dysplasia and other risk factors such as infection by a high risk (oncogenic) strain of HPV.

[0073] The disclosed methods and compositions are primarily intended for risk assessment based upon the evaluation of cytological specimens in a cancer screening environment, but it may also be used with biopsy specimens and for diagnostic and patient management purposes. Touch preps are a particularly convenient form of biopsy specimen for this purpose, but other forms such as tissue slices and dispersed tissues may also be used. A cervical cytology sample such as is routinely used in cervical cancer screening will be assumed in the description below.

[0074] 2. A Method of Predicting Risk that Dysplasia Will Progress to Cancer or that a Cancer Will Progress.

[0075] A method of predicting risk that dysplasia will progress to cancer or that a cancer will progress includes:

[0076] detecting cytokines;

[0077] identifying dysplasia and cancerous cells; and

[0078] correlating the results of (a) and (b) into a classification used for risk evaluations.

[0079] A morphological stain is used in the detection of dysplastic and cancerous cells and immunological reagents for the detection of cytokines. Standard chromatic stains and chromatically labeled immunological reagents may be used for this purpose, but a fluorescent nuclear or morphological stain in combination with fluorescently-labeled immunological reagents is more convenient for most purposes. In particular, fluorescent staining facilitates concurrent or sequential staining of the same cell for multiple analytes whereas chromatic staining often necessitates applying each stain to a different specimen prepared from the same sample. Performing all staining on the same specimen greatly simplifies the correlation step and produces substantially more robust results.

[0080] A standard two step immunological staining procedure using a labeled secondary antibody is described, but other formats such as one using a labeled primary antibody may also be employed. Similarly, although this method may be performed using standard visual microscopy techniques, automated image capture in combination with automated analysis of these captured images is more convenient and effective, particularly when large numbers of specimens, such as are encountered in screening programs, must be evaluated.

Definitions

[0081] Screening as defined herein is the examination of a sample with the goal of detecting the presence of an abnormality. This translates to the detection of cells that are dysplastic or cancerous. The operative word in this definition is detecting. It says nothing about determining the specific type or cause of the dysplasia or what it signifies. A lot of the confusion surrounding this term stems from the common practice of also reporting a degree of dysplasia (none, low, medium, high) as part of the screening results.

[0082] Diagnosis, on the other hand, as defined herein has nothing to do with detection, but rather it focuses upon the detailed classification and the identification of the underlying cause of an already detected abnormality according to, for example, specific type, degree and other characteristics. As an example, a breast cancer screening result might be reported as atypical hyperplasia (the equivalent to high grade dysplasia in some other tissues) whereas the corresponding diagnosis may be might be stage 3 invasive ductal epithelial carcinoma further characterized as being negative or positive for a variety of cell surface markers such as ER, PR and Her2Neu (absence of all three of these indicates that the diagnosed cancer is of the infamous triple negative type).

EXAMPLES

[0083] The methods include the following steps. Certain of these steps may be omitted, combined, or performed in an order other than as presented depending upon the requirements of any particular use of this method.

[0084] In some cases it may be desirable to supplement this invention by the addition of steps in which the specimen is stained and evaluated for the presence of markers other than those included within the scope of this invention:

[0085] 1. Obtaining a Cellular Sample.

[0086] Numerous cell collection devices and methods have been devised and their uses for the collection of cellular samples from tissues, body secretions, and other anatomical sources are well known in the art. Due to the diversity of such devices and methods that are available and because each such device or method is often intended for the collection of cells from some specific anatomical site, a detailed discussion of such devices and methods is beyond the scope of this description.

[0087] Any such device or method that is capable of collecting a sufficient number of the target cells may be used in the practice of this invention. Between 1000 and 5000 cells is generally sufficient, but a greater number is preferable particularly in cases where target cells are expected to comprise only a small subset of the total cells collected and/or when a significant loss of cells is anticipated during the preparation of a specimen from the sample.

[0088] Standard statistical methods may also be used to estimate the number of cells required to achieve a particular level of statistical confidence in the test result given an approximate value for the incidence rate of the disease state of interest in the population being tested. In addition to providing a sufficient number of target cells, the collection device and method preferably causes little if any damage to the collected cells, minimizes the collection of potentially interfering substances such as mucus and red blood cells, and satisfies various operational criteria related to ease of use, consistency, invasiveness, cost and similar factors. The target cells that are suitable for use in the practice of this invention are of epithelial origin. As available collection devices and methods are generally not highly selective as to the type of cell collected, non-target cells, which may be of epithelial or non-epithelial origin, are also present in the collected sample. The class of non-target cells that is of relevance to the practice of this invention are lymphocytes, comprising T-cells, B-cells, neutrophils, macrophages, myeloid-derived suppressor cells and other cells of the immune system.

[0089] In one preferred embodiment cervical epithelial cells may be collected using a cervical spatula or broom to scrape or abrade such cells from the surface of the cervix. In another preferred embodiment, a vacuum device is employed to cause the milk ducts in breast tissue to express a fluid (nipple aspirate fluid) containing ductal epithelial cells that are collected by absorption onto an absorbent material, filtration through a membrane filter, or by sedimentation. Similarly, in other preferred, but not limiting embodiments, bladder cells may be collected from urine or bladder washings by filtration or sedimentation; lung cells may be collected from sputum; skin cells may be collected using any of a variety of scraping or adhesive devices and cells from tissues that are not directly accessible externally may be collected by fine needle aspiration (FNA). In some cases tissue samples obtained incidentally to a surgical procedure may be available. In such cases cells may be recovered from the tissue by dispersion of the tissue in a fluid medium or preferably by the touch prep method in which the surface of the tissue is brought into momentary contact with the surface of a microscope slide that has previously been coated with a material such as poly-L-lysine to which cells preferentially adhere. The samples collected by these and by other similar methods comprise mixtures of target and non-target cells.

[0090] Because collected cells are generally unstable and there is a potentially significant time delay between when the cells are collected and when they are prepared as specimens for evaluation, it is preferred that the freshly collected cells be preserved or fixed immediately after collection. Numerous formulations of preservatives and fixatives and equally numerous methods for their use are well known in the art. For the purposes of this invention the preferred preservative or fixative is alcohol-based or may alternatively be an organic solvent such as acetone or ether. Suitable alcohols include methanol, ethanol and iso-propanol either neat or more commonly in a buffered aqueous solution in which the alcohol concentration is between 15 and 80%. Suitable alcohol-based preservatives may additionally contain mucolytic and other modifying agents while alcohol-based fixatives often additionally contain a polymer such as Carbowax or poly-ethyleneglycol. Under some circumstances a crosslinking fixative such as ones comprising formaldehyde, formalin or glutaraldehyde may be used in the practice of this invention, but such fixatives are not preferred as the crosslinking reactions that occur in such fixatives can block, obscure or damage the epitopes that are to be detected in the immunostaining step of this invention.

[0091] 2. Preparing a Cytological Specimen from the Cellular Sample

[0092] The preparation of a cytological specimen requires that the collected cells be transferred to a microscope slide and then stained in a manner that highlights the cellular features and constituents of interest.

[0093] Numerous devices and methods for the deposition of cells onto a microscope slide are well known in the art and may be used in the practice of this invention. The preferred method is one that minimizes the number of cell clumps on the slide, the degree of overlap between adjacent cells on the slide, and the loss of cells and the damage done to cells during the transfer process. One such preferred method is cytocentrifugation in which the cells contained in a cell suspension are deposited onto a microscope slide under the influence of a centrifugal (gravitational) field. One limitation of this method is that contaminants such as mucus and red blood cells that may be present in the cell sample can be deposited on the slide in a manner that obscures target and non-target cells. In such cases the use of a cell preservative (see above) that contains suitable mucolytic and/or other selective lytic agents is appropriate. Alternatively, the removal of unwanted sample constituents by any of a variety of filtration or gradient centrifugation methods is widely practiced in the art. Such separations are an integral part of several widely used methods in which a membrane filter is used to transfer cells to the slide. Methods that utilize or incorporate cell separation may be used in the practice of this invention so long as the separation method does not result in an unacceptable degree of loss of target or non-target cells.

[0094] 3. Treating the Cytological Specimen with One or More Morphological Stains.

[0095] As cells are visually transparent, it is necessary to treat the cells in a manner that establishes a detectable contrast between the cellular structures and other constituents of interest and their surroundings. This contrast is created by staining the cells with various reagents. The stains most commonly used in the evaluation of cytological specimens are chromatic stains such as hematoxylin, which stains DNA, and eosin, which stains various constituents of the cytoplasm. While these stains may be used separately, they are most commonly used in a combination that is often referred to as H&E stain. In some cases thionin is substituted for hematoxylin and/or other stains are combined with H&E for particular purposes. These conventional morphological stains can be used in the practice of this invention if this morphological staining is performed after the immunochemical staining procedure described below and a means such as recording the locations and identities of relevant morphologically identified target and non-target cells is provided in order to allow correlation of these morphological results with the results obtained from the immunochemical staining process described below.

[0096] Practice of the disclosed and claimed methods are greatly facilitated if the chromatic morphological stains such as H&E described above are replaced with fluorescent or fluorogenic morphological stains such as DAPI, POPRO and the like. The advantages of fluorescent staining over chromatic staining, especially when quantitative measurements of this staining are to be made, are well known in the art. Two of these advantages that are particularly pertinent to the present invention are that the wide range of concentrations of cellular constituents present in cells is more readily accommodated in fluorescence and fluorescent and fluorogenic morphological stains are more compatible with fluorescent immunostaining methods than are most chromatic stains. In the context of this last point, it should be noted that the chromatic stain eosin has a fluorescent emission that can overwhelm the emissions from most of the fluorophores commonly used in immunostaining procedures. For these reasons the use of fluorescent or fluorogenic morphological stains is strongly preferred in the practice of this invention. These fluorescent stains and their methods of use are well known in the art.

[0097] In certain cases it may be convenient or desirable to employ a cell-type selective immunological or other type of stain or stains in order to specifically detect target or non-target cells of interest. Examples of such selective staining include use of an immunostain comprising an anti-cytokeratin 19 primary antibody for the detection of ductal epithelial cells in a sample comprising nipple aspirate fluid, or the use of an anti-CD4 or anti-CD8 antibody for the detection and identification of specific classes of T-cells. Such selective staining methods and procedures are well known in the art.

[0098] 4. Detecting, Identifying and Classifying any Cells and Other Objects that May be Present in the Cytological Specimen on the Basis of Morphology and Detecting the Presence of Target Cells Having Characteristics Indicative of Dysplasia, Cancer, Infection or Other Disease State on the Basis of Cell Morphology

[0099] The methods and criteria for the morphological identification and classification of cells from various tissues are well known and established in the art. The identification and classification of target epithelial cells and of non-target cells of the immune system such as T-cells, B-cells, neutrophils, macrophages, and myeloid-derived suppressor cells is performed in the practice of this invention. Other types of cells may generally be ignored. Cell identification and classification is traditionally accomplished by the visual assessment of cell morphology, but it may also be beneficially accomplished by the automated analysis of electronically captured images of cells.

[0100] 5. Treating the Cytological Specimen with One or More Immunological Stains and Detecting the Presence of One or More Cytokines.

[0101] Numerous methods for the immunostaining of cellular specimens for the purpose of detecting and optionally quantitating specific cellular constituents have been developed since this technique was first introduced in the 1950's. Of these, the labeled secondary and ELISA are formats are presently the dominant methods employed in the cytology laboratory. The labeled secondary format is the preferred immunostaining method in the practice of this invention. This format utilizes an unlabeled primary antibody that binds specifically to the cellular constituent of interest in combination with a fluorescently-labeled secondary antibody that binds specifically to the Fe portion of the primary antibody. Formats utilizing fluorescently-labeled primary antibody can also be used effectively, but can result in lower signal levels and can require custom labeling of the antibody.

[0102] Primary antibodies that bind strongly and specifically to the anti-inflammatory cytokines TGF-, IL-4, IL-10, IL-13 and interferon-alpha (INF-), either individually or in combination, are preferred for use in this invention. Similarly, primary antibodies that bind strongly and specifically to the inflammatory cytokines IL-1, IL-12, IL-18, IL-23, tumor necrosis factor (TNF), interferon-gamma (IFN-) and granulocyte-macrophage colony stimulating factor (GMCSF) are preferred for use in this invention. Of the anti-inflammatory cytokines, IL-10 has proven to be the most generally applicable to the detection of dysplasias that are likely to progress in a variety of tissues while the other anti-inflammatory cytokines such as IL-19 appear to be more tissue specific. In an analogous manner, of the inflammatory cytokines, IL-12 and IL-23 are the most generally applicable as they are broadly involved in the activation and differentiation of TILs and natural killer (NK) cells in the microenvironment surrounding a dysplastic lesion. Therefore, the target cytokine or combination of cytokines utilized in the practice of this invention is best determined on the basis of the tissue from which the cell sample to be tested was obtained. The performance of this invention will depend upon the particular antibodies selected. Due to the relatively low concentrations of the cytokines that are present in the samples used in this invention and the relatively low signal amplification available in the labeled secondary assay format, the preferred primary antibodies should have affinities in the low nanomolar, or preferably the low to mid-picomolar range in combination with at least a moderately high specificity for the particular target cytokine.

[0103] The labeled secondary antibody is an antibody raised in a species other than the species of the primary antibody that binds specifically to the Fc portion of the selected primary antibody and to which several fluorophore molecules are attached. Suitable fluorescently labeled secondary antibodies are commercially available from numerous sources or may be prepared by methods that are well known in the art. A disadvantage of fluorophore-labeled secondary antibodies is that they exhibit a certain amount of nonspecific binding, which makes rare event detection problematic. In the present invention this is overcome by the use of secondary antibodies labeled with fluorophores of different structural classes and hence different non-specific binding properties. Examples of such dyes are Alexa Fluor 594 and Alexa Fluor 647

[0104] Immunostaining of the specimen is carried out in the manner well known in the art by treating the specimen with the selected primary monoclonal antibody or antibodies specific for the target anti-inflammatory cytokines; removing unbound primary antibody by washing; treating the specimen with a mixture consisting of approximately equal parts of secondary antibody labeled with the first and second fluorophores; removing unbound secondary antibody by washing; and protecting the stained slide by the application of a coverslip.

[0105] The results of the immunostaining are assessed by examination of the stained specimen under a fluorescence microscope that has been configured to utilize excitation and emission wavelengths that are compatible with the first and second fluorophores employed. This assessment may be made visually or preferably by means of an automated image capture and analysis system. The intent of this assessment is to identify those cells, if any, present in the specimen where staining by the antibody labeled with the first fluorophore spatially coincides with staining by the antibody labeled with the second fluorophore. This procedure permits detection of and compensation for the non-specific binding of the labeled secondary antibody to the cells, which if uncorrected could yield false positive results. In general terms, if only one or the other of the labeled secondary antibodies is bound at a particular location on a cell it can be assumed that this binding is non-specific, but that if both labeled antibodies bind at the same location, the binding can be assumed to be specific and therefore a true indicator of the presence of the target anti-inflammatory cytokine.

[0106] 6. Assessing the Relative Risk of the Target Cells Thus Identified Progressing to a More Adverse Disease State Based Upon the Detection of the Presence of One or More Cytokines in Some Portion of the Cells Comprising the Cytological Specimens.

[0107] In order to estimate the risk of progression it is necessary to classify each target and non-target cell, in which the immunostaining procedure of step 5 indicated that the target cytokine(s) is (are) expressed, in terms of its type and, if applicable, its degree of morphological abnormality. As immunostaining and morphological staining are performed on the same specimen, the immunostained cells detected in step 5 can be identified and classified based upon the morphological information pertaining to the same cell that was obtained in steps 3 and 4. This correlation may be performed manually, but is most conveniently performed using an automated image analysis system. It is also useful, but not necessary in the practice of this invention, to identify any unstained morphologically abnormal target cells and any unstained cells of the immune system that may be present.

[0108] The risk of progression can initially be stratified by morphologically identifying the cells comprising the sample and, for each such cell, determining whether the cell is expressing IL-10 and/or other anti-inflammatory cytokines. Knowledge of the type of cell in combination with whether the cell is expressing one or more anti-inflammatory cytokines allows estimation of whether the immune response to an abnormal cell has occurred; whether the immune response to such an abnormality has terminated normally upon resolution of the abnormality; whether the immune response has been terminated prematurely; and, if premature termination has occurred, whether this termination has occurred through normal processes or has been forced by anti-inflammatory cytokines being secreted by abnormal target cells. The type(s) of cell(s) expressing one or more anti-inflammatory cytokines allows assessment of the risk of progression.

[0109] The of the risk of progression estimated in accordance with the preceding paragraph can be further refined by additionally considering the expression of inflammatory cytokines such as IL-12 and IL-23 by these same cells. In particular, inflammatory cytokines such as IL-12 and IL-23 counteract the effects of anti-inflammatory cytokines in a cell type specific manner and that the ratio of the number of cells expressing IL-12 (or IL-23) to the number of cells expressing IL-10 is informative as to the relative amount of immune suppression or activation of the tumor microenvironment.

[0110] An example of the results obtained by the above procedure is summarized in Table 1. Each sample was collected as part of the routine Papanicolaou (Pap) screening of women for the presence of cervical cancer or during follow up of women having a previous abnormal screening result. These samples were collected from the cervix through the use of a cervical spatula, sometimes in combination with an endocervical brush, and preserved in a commercially available methanol-based preservative (ThinPrep). One portion of each sample was prepared as a cytological specimen; stained using the Pap stain; and morphologically evaluated by several trained cytologists in accordance with internationally accepted standards and criteria. The results from these conventional cytological evaluations of this portion of the sample are recorded in the Cytology column of Table 1, which is organized by increasingly adverse cytological diagnosis, and comprise the official reference diagnoses for these specimens. HPV status is also provided where available as infection with a high risk (HR, oncogenic) strain of the HPV virus is widely considered to be a primary cause of cervical cancer. Infection or co-infection by low risk (LR, non-oncogenic) strains of HPV are also noted when available.

[0111] A second portion of each sample comprising approximately 5000 total cells was prepared by cytocentrifugation; stained using PO-PRO; immunostained using a rabbit monoclonal anti-IL-10 primary antibody in combination with goat-anti-rabbit secondary antibodies labeled with Alexa Fluor 594 and Alexa Fluor 647. Morphological evaluation, data capture and data analysis were performed both visually using a commercially available fluorescence microscope (Olympus BX-50) and using a custom built automated image capture and analysis system. The correlation between IL-10 staining and morphological classification is shown in the HSIL and HSIL Group columns in Table 1. In this particular study a reporting threshold of HSIL was used as this level of morphological abnormality is often considered to be the threshold for initiating aggressive patient follow up. HSIL and HSIL Groups are reported separately as HSIL Groups are often considered to be a more adverse result than is HSIL in isolated cells. The Risk Assessment column identifies the morphological classification(s) of any IL-10 expressing squamous epithelial cells and the corresponding estimated level of risk that the patient was likely to progress to a more adverse disease state than indicated in the Cytology column.

[0112] Sample 10 in Table 1 indicates the ability of this invention to detect the presence of abnormal cells in a nominally normal sample. In this specific instance both IL-10 positive isolated HSIL and a clump of IL-10 positive HSIL were found. Thus although this sample was reported as being normal by conventional cytology, this patient should be considered as being at high risk of progression to cancer.

[0113] Sample 15 in Table 1 is classified as ASCUS by conventional cytology. ASCUS is a category that is used predominantly in the US to denote the presence of squamous cells exhibiting morphological abnormalities that do not fall within the traditional categories of LSIL, HSIL or Cancer. It is thought that at least some fraction of samples reported as ASCUS represent sample collection errors in which the sample was collected from an area adjacent to, rather than including, a lesion that if properly sampled would return a diagnosis of LSIL or higher. As a significant portion of the abnormal results reported during the Pap screening for cervical cancer are classified as ASCUS, a method for identifying that subset of ASCUS samples that are clinically significant is desirable. This particular sample was found to contain IL-10 expressing cells that upon morphological evaluation were found to be HSIL or HSIL clumps and that the patient should therefore be considered to be at a high risk of progression.

[0114] Under the Bethesda System for the classification of cells in cervical cancer screening LSIL (samples 16-42 in Table 1) is a heterogeneous category that includes both cells exhibiting the morphological changes associated with low grade dysplasia and cells that exhibit the morphological changes associated with infection of cells by viruses such as HPV. It is well known in the art that a significant fraction, over 90% by many estimates, of cells classified as LSIL will spontaneously revert to normal over time while the remainder will progress to HSIL. As is the case for ASCUS, these considerations indicate that a means of identifying that subset of LSIL samples that are likely to progress is desirable. LSIL samples 16-34 in Table 1 do not contain IL-10 expressing squamous epithelial cells and are therefore not considered to be at an increased risk of progression while LSIL samples 35-42 are positive for IL-10 progression and were found to contain HSIL and/or HSIL groups upon reexamination. It is of interest to note that cells that were morphologically classified as LSIL did not express IL-10. Another factor of note is that some of these LSIL samples, including sample 41 which was judged to be at high risk of progression, tested negative for the presence of High Risk HPV infection. This type of observation has implications for the proposed use of HPV testing as the primary means of cervical cancer screening.

[0115] As is the case for LSIL, a certain portion of cases classified as HSIL are known to spontaneously revert to normal while the remaining cases progress to cancer. This portion of cases that spontaneously regress has been variously estimated as being between 30 and 60%. The lack of IL-10 expression reported for samples 43-47 suggests that these samples are likely to regress whereas samples 48-63 express IL-10 and are likely to progress if not treated.

[0116] Table 2 provides an example of the results obtained when the expression of both anti-inflammatory (IL-10) and inflammatory (IL-12) cytokines are assessed in combination with cell morphology in cervical cytology samples. These samples were obtained, processed and assessed in the same manner as described in [00068]-[00069] except that expression of both IL-10 and IL-12, rather than IL-10 alone, was assessed. In Table 2, the columns headed IL-12 Score and IL-10 Score reflect the number of cells expressing the indicated cytokine normalized to a total of 10,000 cells in the sample. The column headed Immune Score reflects the total number of cytokine-expressing cells, again normalized to a total of 10,000 cells in the sample, while Immune Strength is a qualitative representation of the cellular immune activity in the sample. The column headed IL-12/IL-10 Ratio is the ratio of the cell counts in the corresponding Score columns, these ratios being described qualitatively n the Immune Status column and translated into a qualitative Risk of Progression in the final column.

[0117] In addition to the embodiment described above, steps may vary depending upon the requirements of a specific application. For example, the order of steps 4/5 and step 6 may be reversed or, is a fluorescent or fluorogenic morphological stain is used, these steps can be combined into a single step. Similarly, morphological staining and evaluation can be performed only on those specimens on which evidence of immunostaining for the presence of anti-inflammatory cytokines are noted or morphological staining and evaluation can be deleted entirely and the presence or absence of staining for an anti-inflammatory cytokine can be taken as the indicator of whether additional follow up is required. This invention can also be extended by performing staining and evaluation for the presence of additional markers such as PD-1, PD-L1, CTLA-4, p-16, and Ki-67 in order to obtain additional information about a particular specimen.

[0118] 7. Obtain a Suitable Sample.

[0119] In the case of cervical cancer screening, this sample traditionally comprises primarily epithelial cells that are obtained from the cervix by scraping or brushing, but also contains other cell types that are present in the tissue sampled.

[0120] 8. Prepare a Cytological Specimen from the Sample.

[0121] Cytology specimens for cervical cancer screening are traditionally prepared by using the cell collection device to physically smear the collected cells onto a microscope slide. Specimens prepared in this manner may be used with the caveat that some cells of interest may be embedded in clumps or clusters or otherwise obscured and therefore not readily evaluated.

[0122] The monolayer or liquid based method of specimen preparation was introduced in large measure to address the obscuration issues inherent in smear type preparations. In these methods the collected cells are washed from the cell collection device into a liquid medium thereby forming a cell suspension. The cells contained in this suspension are then deposited onto a microscope slide by any of several established means.

[0123] Capture of the suspended cells on the surface of a membrane filter, followed by either placing the filter with captured cells on a microscope slide, or transferring the captured cells from the filter to a microscope slide, includes one of the two major classes of such methods that are in widespread use. Specimens prepared in this manner can be used provided that a suitable filter is used. The filters most commonly used in preparing specimens for cervical cancer screening are designed to capture, on the surface of the filter, the epithelial cells that are traditionally evaluated, while allowing other cell types that may be present in the sample, including T-cells, macrophages and other types employed in the practice of this invention, to pass through the filter and into a waste container. Unless a filter having smaller pores and being capable of capturing T-cells, macrophages and other types of cells employed in the practice of this method is used, the numbers of these cells in the resulting specimens will be significantly reduced relative to their numbers in the original sample.

[0124] The other commonly practiced method of preparing specimens from cell suspensions relies upon the settling of the cells in the suspension onto a microscope slide in the presence of a gravitational field. Methods in which this settling occurs in a 1G field and methods in which this field is increased to several hundred G by, for example by centrifugation (e.g., a CytoSpin) are commonly employed and may be used. However, the specific version of the settling method that has received regulatory approval for use in the preparation of specimens for cervical cancer screening requires that the sample be separated into an epithelial cell fraction and a fraction containing the other types of cells present in the sample, and that only the epithelial cell fraction be deposited onto the slide for cytological evaluation. As in the case of the filtration method described above, the numbers of T-cells, macrophages and other types of cells that are present on a slide prepared by this approved method will be significantly reduced relative to their numbers in the original sample. For this reason it is preferred in the practice of this invention that fractionation of the sample prior to deposition be omitted. Alternatively, the fractions containing the epithelial and other types of cells can be separately recovered and evaluated.

[0125] 9. Detect and Optionally Classify any Dysplastic and/or Cancerous Cells that May be Present in the Specimen.

[0126] The morphology based cytological detection and classification of dysplastic and cancerous cells is carried out in accordance with internationally accepted practices and standards. In these practices detection and classification is facilitated by staining the specimen prior to examination. The Pap stain is commonly used in the screening of cervical cytology specimens while H&E with or without various counterstains is typically used with other types of specimens. These and other stains may be used.

[0127] The cytology classification reported traditionally applies to the entire specimen. A specimen level result can be used, but it is preferable that each cell detected in the specimen be individually classified. An automated image capture and analysis system is a convenient means of providing this function.

[0128] The results of this classification are reported in accordance with any of several standardized systems of nomenclature. The major reporting categories in the widely used Bethesda nomenclature system and corresponding follow up practices are:

[0129] Within Normal Limits (WNL, typically also includes reactive and repairative) [0130] No follow up or intervention warranted.

[0131] Low grade Dysplasia (LSIL)

[0132] Between 90 and 95% of all LSIL spontaneously revert to WNL as the lesion is cleared by the immune system. Local standards of care generally do not recommend follow up other than a shortened interval between screenings unless other risk factors are present.

[0133] High Grade Dysplasia (HSIL)

[0134] Standards of care dictate therapeutic intervention if HSIL is confirmed by biopsy. Between 60 and 90% of all HSIL are spontaneously cleared by the immune system with the balance progressing to cancer.

[0135] Cancer (CA)

[0136] Immediate therapeutic intervention if CA is confirmed by biopsy.

[0137] Atypical Squamous Cells of Undetermined Significance (ASCUSrecognized in some countries)

[0138] The large majority of ASCUS findings are determined to be benign on follow up, but a small percentage are suggestive or indicative of HSIL or CA. Follow up is generally determined on a case-by-case basis.

[0139] As can be seen from the foregoing, although the present invention can be applied to specimens in any of these categories, it is most beneficial when applied to specimens that are classified as being HSIL or ASCUS on the basis of morphology and to LSIL specimens if other risk factors are present.

[0140] 10. Detect the Presence of Inflammatory and/or Anti-Inflammatory Cytokines in any of the Dysplastic Cells, Regulatory T-Cells, Regulatory B-Cells, Myeloid-Derived Suppressor Cells and M2 Macrophages in the Specimen.

[0141] Interleukin 10 (IL-10) is the preferred anti-inflammatory cytokine for the present method because IL-10 plays a central role in terminating the local immune response and has been shown, at least in some cancers, to be involved in other progression events such as suppression of apoptosis and reduction of the display of tumor antigens on the MHC-I complex. Other anti-inflammatory cytokines such as TNFmay also be used. IL-12 is the preferred anti-inflammatory cytokine.

[0142] Detection of these cytokines is most conveniently accomplished by immunostaining. Any standard method of immunostaining can be used. The preferred immunostaining format utilizes an unlabeled primary antibody in combination with a fluorescently-labeled secondary antibody. Formats utilizing fluorescently-labeled primary antibody can also be used effectively, but can result in lower signal levels and can require custom labeling of the antibody.

[0143] 11. Assessing the Risk that the Dysplasia in the Patient from Whom the Specimen was Obtained Will Progress to Cancer or that a Cancer is Likely to Progress.

[0144] The risk of progression of dysplasia or cancer is in large measure controlled by the microenvironment surrounding the lesion. This tumor microenvironment (TME) can be tumor suppressive (inflammatory), permissive, or tumor promoting (anti-inflammatory) depending upon the relative expressions of inflammatory and anti-inflammatory cytokines by both immune cells and cancer cells within this environment. Both the type(s) of cytokine(s) expressed and the type(s) of cell(s) expressing the specific cytokine(s) are significant in assessing the risk of progression. In this assessment:

[0145] The cell type and, where appropriate, whether the cell is dysplastic or otherwise abnormal, is determined from the morphology of the cell,

[0146] The inflammatory tendency of the cell is assessed by determining whether the cell is expressing one or more inflammatory cytokines.

[0147] The anti-inflammatory tendency of the cell is assessed by determining whether the cell is expressing one or more anti-inflammatory cytokines.

[0148] The immunological nature of the TME is determined by the relative numbers of cells in the TME that are expressing inflammatory and anti-inflammatory cytokines, as modified by the specific type of cell expressing the cytokine(s). An immune cell expressing an anti-inflammatory cytokine, for example, has a different significance than a morphologically abnormal (e.g., dysplastic) cell expressing the same anti-inflammatory cytokine.

[0149] It is assumed that the specimens evaluated by this method predominately includes HSIL plus selected subsets of LSIL and ASCUS in which the patient has additional recognized classical risk factors such as a family or personal history of cancer, DES exposure, or a persistent infection by an oncogenic virus (such as a high risk strain of HPV).

[0150] The presence of an anti-inflammatory cytokine, such as IL-10, in any cell in the specimen can be taken as evidence that the local immune response is being, or has been terminated. If expression of an anti-inflammatory cytokine such as IL-10 is not detected in the dysplastic cells in the specimen, it is reasonable to assume that this termination is a normal response to prolonged stimulation of the immune system. As the dysplasia is still present, there is a moderate risk of progression.

[0151] The presence of dysplastic cells expressing an anti-inflammatory cytokine such as IL-10 can be taken as evidence that the local immune response is being actively suppressed and that the patient is therefore at a high risk of progressing to cancer, especially if regulatory cells are also present.

DISCUSSION

[0152] It is not unusual for cells expressing an anti-inflammatory cytokine such as IL-10 to be rare, even in specimens that are classified as HSIL and cancer on the basis of morphology. This is largely due to the use of the traditional cell collection methods which generally collect T-cells, macrophages and other cells of the immune system only incidentally to the collection of the targeted cell type. Because it is often desirable, if not necessary, to employ samples collected using these traditional methods to produce the method disclosed and claimed herein, it is necessary to optimize the immunostaining reagents and methods to maximize detection of the target cell types. A wide variety of optimization techniques are well established in the art and can be used. However, fluorophores such as those used in the practice of this invention are known to have tendencies toward binding non-specifically to cellular constituents in a manner that is not effectively controlled by standard optimization techniques. For this reason, another aspect of the present invention is directed toward mitigating the effects of this non-specific binding on the sensitivity of the test.

[0153] Non-specific binding in immunoassays is addressed by the use of various combinations of blocking agents, detergents, chaotropic agents and other additives. Although these methods are generally effective in limiting or suppressing most types of non-specific binding, their effectiveness in mitigating the non-specific binding of fluorophores to cellular constituents is not adequate for the practice of this invention. The present method therefore uses correlated double staining to address this deficiency.

[0154] The previously described immunostaining format including an unlabeled primary antibody in conjunction with a labeled secondary antibody is used in this description. It is further assumed that the label is a fluorophore selected from the widely used Alexa series. When used in the tests of the present method it can readily be observed that different members of the Alexa series exhibit different patterns of non-specific binding to cellular material. Thus the observed staining pattern includes the superposition of the specific staining of the target analyte and the non-specific binding of the fluorophore to the cellular material. If a different fluorophore is used as a label, the analyte specific staining pattern will remain the same while the pattern due to non-specific binding of the fluorophore will change. This consistency in analyte specific staining in combination with the variability in fluorophore specific non-specific staining provides a method of minimizing the effect of non-specific fluorophore binding on the analytical result.

[0155] Correlated double staining utilizes two aliquots of the same secondary antibody, one of which is labeled with a first fluorophore and the second of which is labeled with a second fluorophore. By way of example, the first aliquot is labeled with Alexa-647, which has a fluorescent emission in the near infra-red spectral region, and the second is labeled with Alexa-594 which has a fluorescent emission in the red spectral region. These two labeled secondary antibodies are combined before use.

[0156] As previously described, the specimen is first treated with an unlabeled primary antibody that binds specifically to the analyte of interest. The specimen is then treated with the combined labeled secondary antibodies and images of the specimen are captured in both the red and infra-red spectral regions. These red and infra-red images are then correlated to identify those regions of the specimen in which the analyte is present.

[0157] If, for a given area, a low signal is detected in both the red and infra-red images, no staining has occurred at the corresponding location in the specimen and the target analyte is not present at that location.

[0158] If, for a given area, a high signal is detected in both the red and infra-red images, and the signals are correlated, the corresponding location in the specimen has been stained selectively and the analyte is present at that location.

[0159] If, for a given area, a high signal is detected in either the red or the infra-red image, but not in both, the corresponding location in the specimen has been stained non-specifically and the analyte is not present at that location.

[0160] FIG. 1 is a scattergram that shows results of IL-10 immunostaining of an IL-10 negative HSIL sample. The detection reagents consist of a primary antibody specific for IL-10 and a mixture consisting of a secondary antibody labeled with Alexa-594 (green fluorescence) and the same secondary antibody labeled with Alexa-647 (red fluorescence) FIG. 1B: Cells exhibiting low spatial correlation (R-G Correlation Coefficient=0.088) between red and green fluorescence thus indicating non-specific binding of the secondary antibodies. The predominantly red fluorescence indicates that these cells non-specifically bind the Alexa-647 labeled secondary antibody preferentially over the secondary antibody labeled with Alexa-594. FIG. 1 A: Morphologically normal and abnormal cells exhibiting high spatial correlation (R-G Correlation Coefficient >0.9) between red and green fluorescence. This indicates that both labeled secondary antibodies are bound specifically. The low fluorescence intensities in this region indicate that little or no IL-10 is present in either morphologically normal (black squares) or dysplastic (red squares). FIG. 1C: Cells exhibiting low spatial correlation (R-G Correlation Coefficient=0.67) between red and green fluorescence thus indicating non-specific binding of the secondary antibodies. The predominantly green fluorescence indicates that these cells non-specifically bind the Alexa-594 labeled secondary antibody preferentially over the secondary antibody labeled with Alexa-647.

[0161] FIG. 2 is a scattergram that shows results of IL-10 immunostaining of an IL-10 positive HSIL sample. Staining of these cells and the interpretation of are as described in the legend for FIG. 1. Cells exhibiting both high red and green fluorescence intensities indicate strong IL-10 production and high spatial correlation (R-G Correlation Coefficient >0.9) between red and green fluorescence indicating that binding of the secondary antibodies is specific.

[0162] FIG. 3 shows color separated images of the same cell showing strong (>0.90) spatial correlation of the localization of the Alexa-594 (green) and Alexa-647 (red) labeled secondary antibodies. Blue denotes staining of the cell nucleus by DAPI. Co-localized red and green emission appears as yellow in this image. Color separated images of the same cell showing strong (>0.90) spatial correlation of the localization of the Alexa-594 (green) and Alexa-647 (red) labeled secondary antibodies. Blue denotes staining of the cell nucleus by DAPI. Co-localized red and green emission appears as yellow in this image. IL-10 negative cells in this image set are indicated by blue nuclei in a green surround (non-specific binding of the Alexa 594 labeled secondary antibody)

TABLE-US-00001 TABLE 1 IL-10 positive IL-10 positive Dysplasias Found and Sample Cytology HPV Status HSILs HSIL Groups Risk Assessment 1 Normal Unknown Negative Negative Normal risk 2 Normal Unknown Negative Negative Normal risk 3 Normal Unknown Negative Negative Normal risk 4 Normal Unknown Negative Negative Normal risk 5 Normal Unknown Negative Negative Normal risk 6 Normal Unknown Negative Negative Normal risk 7 Normal Unknown Negative Negative Normal risk 8 Normal Unknown Negative Negative Normal risk 9 Normal Unknown Negative Negative Normal risk 10 Normal Unknown Positive Positive HSIL. High risk for progression 11 ASCUS Positive Negative Negative Normal risk 12 ASCUS Negative Negative Negative Normal risk 13 ASCUS Negative Negative Negative Normal risk 14 ASCUS Positive Negative Negative Normal risk 15 ASCUS Positive Positive Positive HSIL. High risk for progression 16 LSIL Positive Negative Negative Normal risk 17 LSIL Positive Negative Negative Normal risk 18 LSIL Positive Negative Negative Normal risk 19 LSIL LR Positive Negative Negative Normal risk 20 LSIL Negative Negative Negative Normal risk 21 LSIL HR Positive Negative Negative Normal risk 22 LSIL HR, LR Positive Negative Negative Normal risk 23 LSIL HR, LR Positive Negative Negative Normal risk 24 LSIL LR Positive Negative Negative Normal risk 25 LSIL HR Positive Negative Negative Normal risk 26 LSIL HR Positive Negative Negative Normal risk 27 LSIL HR Positive Negative Negative Normal risk 28 LSIL HR Positive Negative Negative Normal risk 29 LSIL Negative Negative Negative Normal risk 30 LSIL HR Positive Negative Negative Normal risk 31 LSIL HR Positive Negative Negative Normal risk 32 LSIL HR, LR Positive Negative Negative Normal risk 33 LSIL HR Positive Negative Negative Normal risk 34 LSIL HR Positive Negative Negative Normal risk 35 LSIL HR Positive Negative Positive LSIL-HSIL. High risk for progression 36 LSIL Positive Positive Negative LSIL-HSIL. High risk for progression 37 LSIL HR Positive Positive Negative LSIL-HSIL. High risk for progression 38 LSIL HR Positive Positive Negative LSIL-HSIL. High risk for progression 39 LSIL HR, LR Positive Positive Negative LSIL-HSIL. High risk for progression 40 LSIL HR Positive Positive Negative LSIL-HSIL. High risk for progression 41 LSIL Negative Positive Positive LSIL-HSIL. High risk for progression 42 LSIL HR, LR Positive Positive Positive LSIL-HSIL. High risk for progression 43 HSIL HR Positive Negative Negative Moderate risk for progression 44 HSIL HR Positive Negative Negative Moderate risk for progression 45 HSIL HR Positive Negative Negative Moderate risk for progression 46 HSIL HR Positive Negative Negative Moderate risk for progression 47 HSIL HR Positive Negative Negative Moderate risk for progression 48 HSIL Positive Negative Positive High risk for progression 49 HSIL Unknown Negative Positive High risk for progression 50 HSIL HR Positive Negative Positive High risk for progression 51 HSIL Unknown Positive Negative High risk for progression 52 HSIL Unknown Positive Negative High risk for progression 53 HSIL Unknown Positive Negative High risk for progression 54 HSIL Unknown Positive Negative High risk for progression 55 HSIL HR Positive Positive Negative High risk for progression 56 HSIL HR Positive Positive Negative High risk for progression 57 HSIL HR Positive Positive Negative High risk for progression 58 HSIL Positive Positive Positive High risk for progression 59 HSIL Unknown Positive Positive High risk for progression 60 HSIL Unknown Positive Positive High risk for progression 61 HSIL Unknown Positive Positive High risk for progression 62 HSIL HR Positive Positive Positive High risk for progression 63 HSIL HR Positive Positive Positive High risk for progression

TABLE-US-00002 TABLE 2 IL- 12/IL- Sample Sample HPV IL-12 IL-10 Immune Immune 10 Risk of Number Status Status Score Score Score Strength Ratio Immune Status Progression 1N Normal Pos 146 26 172 Moderate 5.62 Moderately Inflammatory 2N Normal Pos 82 64 146 Moderate 1.28 Weakly Inflammatory 4N Normal Neg 169 52 221 Moderate 3.25 Moderately Inflammatory 5N Normal Pos 162 79 241 Moderate 2.05 Moderately Inflammatory 7N Normal Neg 496 42 538 Strong 11.81 Strongly Inflammatory 1A ASCUS Pos 336 200 536 Strong 1.68 Weakly Inflammatory 1L LSIL Pos 125 0 125 Moderate >100 Strongly Inflammatory 2A ASCUS Pos 31 47 78 Weak 0.66 Anti-inflammatory Likely 2L LSIL Neg 0 0 0 None 0 None 3A ASCUS Pos 519 499 1018 Very 1.04 Weakly Inflammatory Strong 3L LSIL Pos 321 24 345 Moderate 13.38 Strongly Inflammatory 4A ASCUS Neg 95 0 95 Weak >100 Strongly Inflammatory 4L LSIL Pos 85 12 97 Weak 7.08 Moderately Inflammatory 5A ASCUS Pos 62 22 84 Weak 2.82 Moderately Inflammatory 5L LSIL Pos 274 449 723 Strong 0.61 Anti-inflammatory Likely 1H HSIL Pos 47 12 59 Weak 3.92 Moderately Inflammatory 2H HSIL Pos 107 1 108 Moderate 107 Strongly Inflammatory 3H HSIL 140 16 156 Moderate 8.75 Moderately Inflammatory 4H HSIL 9 22 31 Weak 0.41 Anti-inflammatory Likely 5H HSIL 25 6 31 Weak 4.17 Moderately Inflammatory 6H HSIL 187 7 194 Moderate 26.71 Strongly Inflammatory 7H HSIL 90 2 92 Weak 45 Strongly Inflammatory 8H HSIL 365 6 371 Strong 60.83 Strongly Inflammatory 9H HSIL 7 60 67 Weak 0.12 Strongly Anti- Very inflammatory Likely 10H HSIL 165 28 193 Moderate 5.89 Moderately Inflammatory *Positive cells per 10,000 cells **IL-12 Score plus IL-10 Score

TABLE-US-00003 TABLE 3 Risk of dysplasia progressing Cell Cytokine Cell Cytokine Class Present Class Present Risk of Progression WNL No IMM No Normal State WNL No IMM Yes Moderate Increased Risk DYSP No IMM No Slightly Increased Risk DYSP No IMM Yes Moderately Increased Risk DYSP Yes IMM Yes Greatly Increased Risk where: WNL = non-dysplastic epithelial cell; DYSP = dysplastic epithelial cell; and IMM = cell of the immune system.