A systematic approach to cytology (Proceedings)

Cytology is an extremely useful tool that can be employed in any practice. Samples can be acquired quickly and easily, using instruments that are readily available in any veterinary practice from the single veterinarian hospital to the largest referral centers. The strengths of cytology lie in the rapidity with which the samples can be obtained, processed and evaluated. In addition, it is an affordable and generally minimally invasive. Using a systematic, step-wise approach, any specimen can be evaluated and useful interpretations derived.

An exhaustive review of sample acquisition is well beyond the scope or purpose of this lecture. That being said, interpretation of the objective material on a slide can be markedly influenced by the source of the sample. I am a clinical pathologist that believes that the person that can offer the most accurate interpretation of a sample is the person who took it (or was present when it was obtained). Confidence in the source of a sample is important, whether reading the slide in house or sending the slides away for a second opinion, an accurate description of sample acquisition should be available, should the person interpreting the smears want it. The interpretation of a slide is a marriage between the objective material contained within the smear(s) and the additional material provided (i.e. history, physical examination, description of the lesion, etc.). Different people use more or less of the additional information; however, it should be available, should it be needed.

The interpretation of a smear begins when the needle is inserted into the lesion. The gross appearance, texture or appearance using diagnostic imaging can be used in conjunction with the findings on the slide. A gritty feel when the needle is inserted can be indicative of mineralization, a finding which should be corroborated in the smears. An initially clear serous effusion or cerebrospinal fluid that turns bloody should be noted to make interpretation possible. Once the sample is obtained, its gross appearance should be noted. A smear that appears to be composed of water droplets that does not dry upon standing is most likely composed of lipid. If this gross appearance is not noted prior to staining it may go undocumented, Depending on stain used, cell lipid may be removed or dissolved and thus not be present to identify. In addition to lipid, the smear should be evaluated for the overall cellularity. Comparing samples to a smear of pure blood can help get an idea of blood contamination vs. cells from tissue. Finally, gross examination of a smear can help direct you to areas of higher cellularity, which tend to be areas of higher yield.

After gross examination of the smear, it can then be stained for initial evaluation. The type of stain is up to the person evaluating the smears; however, a rapid polychromatic stain is generally employed. Automated stainers are available as well. As long as the stain used provides a consistent, contaminant-free product, you should be fine. Unstained smears should be set aside to perform additional "special" stains if required, either by you or the reference lab that the sample is sent to. Gram stain, New Methylene Blue, Acid Fast, PAS, GMS and immunocytochemical stains are all examples of stains that may be indicated based on the findings of the smears stained with a polychromatic stain.

The microscopic examination begins with a low power "fly by". Areas of interest may be examined longer based on your gross evaluation of the smear. Depending on the type of sample, in most cases a smear will not be homogenous through and through. This homogeneity goes for not only an individual slide, but slide to slide as well, even in the case of "mirror" image slides, where one slide is used to make another. I generally reserve 4X in cytology for two purposes. I will use my 4x objective if presented with numerous slides that are heavily hemodilute, such as a splenic aspirate. The diagnosis of a hemangiosarcoma may be possible, if enough neoplastic cells are found and carefully examined. A 4x lens can aid in finding these cells in a reasonable amount of time. Conversely, a 4x objective can be used to fully evaluate smears that are so hypocellular that they appear clean to the unaided eye. Extremely important material, such as a small collection of mast cells could easily be missed if only a gross evaluation or high power exam are used. When located using a low power scan, an intermediate power lens (10 or 20X) can then be used in the areas that have been identified as higher yield. This intermediate power lenses offer a compromise between ground covered and resolution of detail. I find that I can most easily determine the nature of cell arrangements at 20x. This is important for differentiating epithelial cell populations from other, non-cohesive cells. Agglutination, windrowing and aggregation are all easiest to determine at an intermediate power. Finally, I also use the intermediate powers to insure that when I go to higher power to evaluate the fine details of the smear that I have a list of things (cells, background, etc.) to look at, to prevent missing details or populations of cells. High power (50 – 100x) evaluation is reserved for specific details that need to be resolved such as chromatin patterns, nucleolar features, organisms or fine details in the background. Sine very little of the smear is actually evaluated, I tend to use this power sparingly.

As I am collecting the data from the smear, I mentally go through an algorithm that ultimately leads to an interpretation and subsequent comments and suggestions. The first thing that I always determine is the presence or absence of inflammation. If present, I determine the type and general magnitude, as both can play a roll in the final interpretation. If there is marked inflammation present, my evaluation of the other cell populations will be effected.

Suppurative inflammation consists primarily of neutrophils. Many sources site many numbers; however, the majority (often >75%) of the nucleated cells should be neutrophils. Any time a neutrophil is seen in cytology it should be determined if it is degenerate or non-degenerate. When determining degeneration, I put the greatest emphasis on karyolysis, when the nucleus swells and essentially falls apart within an intact cell. When consistently noted, I am certain that a cell is being killed outright. Pyknosis and karyorrhexis are when the nucleus becomes hyperchromic, markedly condensed and fragment into a few to several pieces, respectively. These changes can be observed; however, they can also be seen with programmed cell death, apoptosis. Degenerate neutrophils prompt me to look for cytologically evident agents that kill neutrophils. My top suspect is bacteria and when seen, I try to demonstrate them being phagocytized to insure that they are indeed pathogenic and not a contaminant. Bile, exposure to air and iatrogenically introduced exogenous agents are other examples of things that can cause neutrophil degeneration. Suppurative inflammation without degenerate neutrophils can be seen with things such as GME, steroid-responsive polyarthritis, trauma and immune mediated diseases. It should be kept in mind that bacterial invasion of an otherwise sterile process is always possible, and expected in certain locations such as the colon or oral cavity.

Determining that neutrophils are present due to inflammation and not blood contamination, which is extremely common, I look for things such as: degeneration, the ratio of neutrophils to erythrocytes (1:400 roughly normal) and other nucleated cells, other cells that suggest inflammation and the presence of platelets, which are only present in contaminated samples or peracute hemorrhage.

Granulomatous inflammation is primarily made up of macrophages. The macrophages present include mononuclear foamy and phagocytic macrophages, epithelioid macrophages, and multinucleated giant cells. Epithelioid macrophages have a single, eccentrically placed nuclei set in a moderate amount of smooth blue cytoplasm. These cells are not phagocytic and appear rather quiescent. If the strong presence of neutrophils is present, the term pyogranulomatous can be used. Differential diagnoses are similar for granulomatous and pyogranulomatous inflammation and include: fungal infection, FIP lesions, foreign body reaction, Mycobacterium sp. infections and sterile processes. True histologic granulomas also contain fibroblasts, lymphocytes and plasma cells, all of which will be variably present cytologically.

The presence of >10% eosinophils signifies eosinophilic inflammation. Often seen as a component of other types of inflammation, eosinophilic inflammation can be seen with a variety of etiologies; however, I use the four broad categories of: allergic / hypersensitivity reactions, fungal infections, parasitic infestations and neoplasia. Specific examples include cryptococcal meningitis, flea bite reactions, mast cell tumors, eosinophilic granulomas complex and eosinophilic bronchopneumopathy.

Lymphocytic and plasmacytic types of inflammation are frequently seen together and may also be a component of other types of inflammation. Vaccine reactions, inflammatory bowel disease (IBD) and non-specific antigenic stimuli can all produce lympho-plasmacytic inflammation.

Once the presence or absence of inflammation is determined, evaluation of the other nucleated cell type(s) is then performed. It is crucial to rule out inflammation, as it can induce morphologic changes in cells that can easily be misinterpreted as malignant changes. Another key to compare cells is to make sure that cells from the same cell line (i.e. apples to apples) are being compared to one another. Cytomorphologic features such as cohesive patterns, cytoplasmic characteristics and arrangement in the smear can be helpful; however, I find that the most consistent feature of cells that can be used to indentify likeness is the chromatin pattern.

The four classes of cells / tissue are: epithelial, mesenchymal, round / discrete and the naked nuclei / neuroendocrine pattern. Epithelial cells tend to be large in size and polygonal in shape. They usually exfoliate easily. The key feature to look for to positively identify them is an organized, cohesive arrangement. If rosette / acinar structures are seen, that suggests a glandular origin to the cells. Mesenchymal cells tend to me medium to large in size and a shape that is some variant of spindled. These cells exfoliate poorly. Fusiform, oblong, caudate and cigar-shaped are all examples of typical mesenchymal cell shapes. These cells can form disorganized aggregates, but cannot form cohesive sheets. In addition, they may either contain a pink granulation and/or be entrapped within a smooth pink material, loosely termed matrix. This "matrix" may be collagen, chondroid or osteoid. The discrete cell / round cell class generally contains five members: mast cells, lymphocytes, plasma cells, transmissible venereal tumors (TVT) and canine cutaneous histiocytomas (CCH) in dogs. As a class, these cells are small, have a round nucleus and exfoliate very easily and are generally not difficult to interpret. Mast cells have pale blue cytoplasm that is variably filled with small, dark purple granules. Lymphocytes can be identified by their consistently high nucleus to cytoplasm ratio (N:C) and finely stippled chromatin pattern. Plasma cells have an eccentrically placed nucleus in a small amount of dark blue cytoplasm that often has a perinuclear clear zone, the Golgi. The nuclei of these cells exhibits a coarse to clumped chromatin pattern. TVT cells have an eccentrically placed nucleus is pale blue cytoplasm that consistently has peripheral punctuate vacuoles. Generally the history and physical examination are as important as the microscopic examination. Finally, CCH cells are mundane cells that have a round nucleus set in extremely pale blue cytoplasm. Often the protein in the background gives these cells a halo appearance. Similar to the TVT, the signalment and physical examination are as important for an interpretation as the microscopic findings. The final class of cells is the naked nuclei / neuroendocrine class. These cells are extremely fragile and thus, finding a large number of round nuclei in a pool of cytoplasm is suggestive of these cells. Knowing the source of a sample (i.e. heart based mass, pancreatic nodule, adrenal mass) can be vital to properly identifying and thus interpreting these cells. On occasion, the source of a population of cells cannot be determined.

After the class of cell is determined and the influence of inflammation is ruled out, the criteria of malignancy can them be evaluated. There are dozens of criteria that the cytopathologist applies to each population of cells to try and determine its biologic behavior. Cytologic features include anisocytosis, abnormal vacuolation, cellular pleomorphism and atypical granulation are all examples of cytologic criteria of malignancy. Nuclear features include anisokaryosis, nuclear molding, coarse chromatin patterns, polykaryosis, nuclear pleomorphism, etc. Nucleolar changes are as described for nuclear. The population as a whole is evaluated and if three strong criteria of malignancy are present, the interpretation of a population is more consistent with malignancy, versus a benign neoplasm or a hyperplastic, which are usually not cytologically different.

Once all of the cellular elements have been accounted for, the background is then evaluated. Erythrocytes are present for any number of reasons; however, if erythrophagia, hemosiderin and/or hematoidin are noted, they represent in vivo hemorrhage. Erythroparasites are rarely noted. Other important features in the background include granules, lipid, matrix and platelets.

After all the objective data on the smear is collected, it can then be applied back to the other case materials to derive an interpretation. There are many circumstances where cytology cannot yield a definitive diagnosis. At that time, the smears can be sent to a reference laboratory for a second opinion. Other times a ceiling is hit and biopsy with histologic evaluation must be pursued. Such a time would be when cells are seen that exhibit criteria of malignancy, yet abundant inflammation is present, obfuscating an interpretation.

References

Allen SW. Cytologic Diagnosis of Neoplasia and Perioperative Implementation. The Compendium of Continuing Education. 8(2): 72-80, February 1986.

Alleman AR, Bain PJ. Diagnosing Neoplasia: The Cytologic Criteria for Malignancy. Veterinary Medicine. 95(3): 204-223, March 2000.

Baker R, Lumsden JH. Color Atlas of Cytology of the Dog and Cat. St Louis, Mosby. 2000.

Ballegeer EA, Forrest LJ, Dickinson RM, Schutten MM, Delaney FA, and Young KM. Correlation of ultrasonographic appearance of lesions and cytologic and histologic diagnoses in splenic aspirates from dogs and cats: 32 cases (2002–2005). Journal of the American Veterinary Medical Association. 230(5):690-696, 2007.

Barton CL. Cytologic Diagnosis of Cutaneous Neoplasia: An Algorithmic Approach. The Compendium of Continuing Education. 9(1): 20-33, January 1987.

Campbell TW. Avian Hematology and Cytology. 2nd Edition. Iowa State Press. Ames. 1995.

Cowell RL, Tyler RD. Diagnostic Cytology and Hematology of the Horse. 2nd Edition. St. Louis. Mosby. 2002.

Cowell RL, Tyler RD, Meinkoth JH, DeNicola DB. Diagnostic Cytology and Hematology of the Dog and Cat. 3nd Edition. Mosby. St. Louis. 2008.

Raskin RE, Meyer DJ. Atlas of Canine and Feline Cytology. W.B. Saunders. Philadelphia. 2001.