Cytology for the General Practitioner (Part 2)

In our first blog on cytology, we briefly described the cytomorphologic categories and went through a brief description of the characteristics of each one. In this blog, we’ll get a bit more specific about those characteristics and go over some common differentials to think about when aspirating masses in general practice. Keep in mind, our job is to find those common things. There are some rarer things we might find on cytology, and that is where the pathologist comes in. As such, this is a scratch on the surface of the many different tumors that exist. But maybe we can find those common things!

Epithelial tumors

Epithelial lesions generally come from glandular or parenchymal tissue. On cytology, the biggest distinguishing characteristic about them is that they like to clump together really tightly. Sometimes this will form tight junctions called desmosomes, which you can see on the slide as clear lines between two very close cells. If you happen to find some cells on their own, they will be large and round to polygonal in shape, with distinct cell borders (meaning you can tell distinctly where the cell ends). Their nuclei tend to be round to oval.

Epithelial tumors can occur on many internal organs, such as the liver, kidneys, stomach, lungs, and urinary tract. However, since the skin has many glands associated with it, we can find plenty of epithelial tumors outwardly on our day-to-day exams. A lot of the most common ones are benign, such as infundibular cysts, apocrine cysts, and nodular sebaceous hyperplasia/sebaceous adenoma.

Infundibular cysts can be seen mostly on dogs 4-8 years old as firm to fluctuant, smooth, round, well-circumscribed skin bumps on the dorsum or limbs. These will typically not have much in the way of nucleated cells to see, as most of it will be amorphous debris and keratinized squamous epithelium or keratin bars. You may also see cholesterol crystals as evidence of cellular degradation within the cyst, which appear as negative-stained rectangles with irregular notches.

Apocrine cysts will also be relatively acellular with only “cystic fluid,” but you may see some of the cuboidal to low columnar epithelial cells that make up the lining (which will often have the above-mentioned tight junctions!).

Sebaceous gland tumors are also super common. It’s hard to tell nodular sebaceous hyperplasia and sebaceous adenoma apart and they both generally act quite benign, so for our purposes we will lump them. They tend to appear in older dogs and cats as a hairless, pink, firm, raised dermal mass resembling a wart (i.e. cauliflower-like). On cytology, you will typically see packs of sebaceous epithelial cells, which look cool because their cytoplasms are very foamy and they have a small, dense central nucleus.

Viral papillomas look like any “old dog wart” grossly, and occur on the head, limbs, and digits. Cytology will show mature keratinocytes with plenty of blue cytoplasm, small round to oval nuclei, and minimal atypia. However, they may have some small bubbles in their cytoplasm, which would be consistent with ballooning degeneration of squamous cells. They may otherwise carry some blue-green keratin deposits or have eosinophilic stippling. The viral ones usually regress within 2 months. However, there are also squamous papillomas (generally just warts) that do not resolve spontaneously.

Squamous cell carcinomas are common in cats and can be quite malignant. They’ll usually grossly look pretty ugly, often ulcerative. They can occur as one or multiple masses, and in cats they tend to be on the face. When looking at cytology, keep in mind that if it’s ulcerated it could also get infected, so there will likely be some purulent inflammation along with the tumor itself. The main finding will be immature, dysplastic squamous epithelial cells. The tumors can be variably differentiated, and their degree of differentiation can make them look slightly different. The more well-differentiated tumors will have more squames and keratinized angular squamous epithelial cells, but with some nuclear atypia. They also may have punctate vacuoles around the nucleus, and they may exhibit emperipolesis, which is when another type of cell passes through the tumor cell (i.e. you may see a neutrophil within a squamous epithelial cell). A more poorly differentiated tumor will have more round, individualized cells with high N:C ratios (the classic “fried egg” appearance). These tumors will also have large amounts of anisocytosis and anisokaryosis, with coarser chromatin and prominent nucleoli.

Trichoblastomas are technically a form of “basilar cell tumor,” but some work has recently been done to more specifically define basilar cell tumors with histology, and it has been determined that most of the common, benign ones we see are likely trichoblastomas, with a hair germ origin. They are usually small, firm intradermal masses, which sometimes ulcerate. On cytology, you can see large clusters of tightly packed basal epithelial cells with large nuclei and very little basophilic cytoplasm. Their nuclei should show very minimal pleomorphism. You can sometimes see granules in these cells (keratohyalin), and you can also see melanin if the mass is in pigmented skin.

Perianal gland adenomas are “the good kind of butt mass.” They can be found near the anus but can also be on the tail or perineum, or even on the prepuce, inner thigh, or midline. They are usually smooth, round, and raised, but can become lobulated and ulcerate as they grow. You’ll typically see sheets of large round cells with a lot of pinkish-blue, finely granular cytoplasm, with round nuclei that may normally have one or multiple prominent nucleoli. This happens to be very similar to what hepatic cells look like, so these are also affectionately called “hepatoid tumors.” They are generally benign, but as you may have experienced, they can still cause issues because of their location. Additionally, perianal gland carcinomas do exist, though they are rare.

Apocrine gland anal sac adenocarcinomas appear as firm subcutaneous masses fixed to the anal sac. On cytology you will see clusters or loosely cohesive sheets of cells with poorly defined cytoplasms, often resembling a naked nuclei appearance. They will tend to be quite pleomorphic, with a high N:C ratio, and may have some small cytoplasmic vacuoles. A key characteristic may be the formation of acinar or rosette patterns, which can help you be more certain that it is an AGASACA and not a hepatoid tumor. Of course, this tumor can be very malignant. Hypercalcemia on a chemistry panel is present in a large majority of cases.

Neuroendocrine tumors

I like to talk about these right after epithelial tumors, because most of them can seem pretty similar to an epithelial morphology. However, many of these will actually present in a naked nuclei morphology, due to the cytoplasm breaking away and the nuclei being left to float around. They will actually tend to give up plenty of cells in your sample and will present as loosely attached sheets of those free nuclei. If you find an intact cell, it will likely have a round to polygonal shape. The nuclei themselves are usually round, though they can be indented, and usually don’t have much in the way of anisokaryosis. As the name implies, these will originate from neural or endocrine tissues. Most of these tumors are quite rare, so likely not as relevant for the scope of this discussion.

Thyroid gland tumors are very complex, so we likely won’t be doing much as far as differentiating types of tumors from thyroid glands. However, if you happen to be poking something that might be coming from the thyroid, you can at least tell if it actually came from the thyroid. Typically, you’ll see small sheets of closely or loosely attached cells, which may contain black, granular material (which is likely tyrosine) in their cytoplasms. Usually, it will appear as the naked nuclei morphology, though sometimes the cells will remain intact with distinct cell borders and scant basophilic cytoplasms.

Mesenchymal tumors

Tumors of mesenchymal cytomorphology will tend to exfoliate individually, though some of them like to create an extracellular matrix, so you may see them grouped up if they are bound together with that matrix. Samples from these are often not very cellular. The cells are typically spindle to oval shaped, but can also be stellate (meaning star-like) or fusiform. Their cytoplasmic borders are often indistinct. The cells are generally not as big as epithelial cells, and their nuclei are usually round to elliptical.

Fibrosarcomas are actually quite common, especially in cats. They tend to live on the limbs, trunk, and head, and are often poorly circumscribed and may be ulcerated. They are malignant. On cytology they will have large, plump cells, often with wispy cytoplasmic tails, presenting individually or grouped together by pink collagenous material. They can have quite marked nuclear pleomorphism, and multinucleated giant cells can also be involved. They can have quite a bit of anisocytosis, as well as a high N:C ratio.

Perivascular wall tumors are decently common in dogs. Grossly they are soft or firm, multilobulated, and well-circumscribed. They’ll usually give up a fair number of cells. You can see plump spindle cells individually or arranged in bundles, with basophilic cytoplasm and often with numerous small vacuoles and sometimes with eosinophilic globules. Their nuclei are ovoid, usually with at least one prominent nucleolus. They can stick to the surfaces of capillaries as well. Some very classic arrangements they can make include a “fingerprint whorl” pattern, as well as a “crown” cell, which is a multinucleated cell with nuclei arranged around an empty (basophilic) center.

Hemangiomas are common and benign. The slide may just look like you only got blood. Sometimes you can see small basophilic endothelial cells. But really, you should look for evidence of hemorrhage within the mass, which would be erythrophagocytosis by macrophages, or macrophages with hemosiderin. Of course, hemangiosarcoma is the malignant blood vessel-associated tumor, but that one is not very common in the skin, and you’d be more likely to run into it if you were poking a spleen or liver, which we don’t do as routinely in general practice.

Melanocytic tumors can be benign or malignant – not all tumors of pigmented skin are melanoma, they can be melanocytomas too. Melanocytomas will have spindle to round-shaped cells with abundant melanin granules which are black-brown or green and uniform in size and shape. Their nuclei are small, without any significant pleomorphism. Malignant melanomas will grossly be infiltrative and often inflamed and ulcerated, but may or may not be well pigmented. Melanoma is a special tumor, because its cells can look like just about anything – round, epithelial, or mesenchymal/spindle. Their nuclei are round to ovoid with often marked anisocytosis, anisokaryosis, coarse chromatin, and prominent nucleoli. They can have lots of melanin, but they also may not have much, which also makes them tricky. Sometimes you may see some cells with a grey, dust-like appearance in the cytoplasm. Amelanotic melanomas are most frequent in the mouth, and since the cells themselves are so wonky, definitive diagnosis more often requires biopsy.

Lipomas, everyone’s favorite! In case you didn’t know (I sure didn’t), they are actually mesenchymal tumors. You will see oily fat on the slide, which will never dry, and when you try to stain it, it will often wash away. I’ve had success actually seeing lipomas by using the hair dryer on low (no heat) from about a foot away, then leaving it in the fixative for at least a minute before the stains. Also, changing the stain (really the fixative) frequently is the key! You can see those big lipocytes with clear cytoplasm and a small, condensed nucleus in the periphery of the cell.

Soft tissue sarcoma is worth mentioning because it isn’t a distinct tumor itself, but technically a term for the collection of mesenchymal neoplasms arising from the soft tissues. An STS on cytology, for example, could be found to be a perivascular wall tumor or peripheral nerve sheath tumor histopathologically. If you poked some nondescript mass and you’re certain it’s a mesenchymal tumor but not certain of the specific type, calling it a soft tissue sarcoma is a perfectly fine start, knowing that biopsy is likely required for the final answer.

Round cell tumors

Round cell tumors will typically exfoliate at least a moderate number of cells, which will be round, have distinct cytoplasmic borders, and present individually. Their nuclei are usually round or indented. Luckily, there are really only five neoplasms in this category, and they can be remembered with the acronym T-LyMPH.

Transmissible venereal tumors are more often seen in free-roaming sexually active dogs, so practicing in suburban parts of the US, I have not run into these personally. Generally, it will be a pink, ulcerated, friable mass, usually bleeding and necrosing, on the genitalia or mucous membranes. It exfoliates readily, often well enough to see it with an impression smear. There are large round cells with large round nuclei, which have coarse chromatin and prominent nucleoli. The cytoplasm is lightly basophilic and has lots of punctate vacuoles.

Lymphoma is not a very common skin lesion, but we of course poke lymph nodes all the time. Interpretation of lymph node aspirates is a very challenging topic that needs an entirely separate blog, written by someone with more expertise on the subject. However, we can briefly go over some general characteristics of lymphoma itself. The biggest red flag would be finding a very homogeneous population of lymphocytes which may have criteria of malignancy like increased mitotic count. It is generally easier to tell that you have an actual lymphoid neoplasm when you see an increased number of immature large lymphocytes, but of course small cell lymphoma is also possible, so the lack of large lymphocytes doesn’t necessarily mean no lymphoma. Comparing the nucleus to the size of a red blood cell can help you categorize small (1-1.5 x RBC), medium (2-2.5 x RBC), or large (>3 x RBC). You may also see an increased amount of lymphoglandular bodies, which appear as small round basophilic blebs, and are the broken fragments of cytoplasm from ruptured cells. They can be seen without cancer but may be increased because neoplastic cells are more fragile.

Mast cell tumors are very common on the skin of both dogs and cats.  They tend to be hairless, raised, and red, but can also be in the subcutis surrounded by fat. I find they like to bleed a lot, so the sample may be hemodilute. The slide will have large, discrete round cells which have the characteristic purple or magenta granules in their cytoplasm and in the background. Eosinophils like to get involved when there is a mast cell tumor, so seeing increased eosinophils may aid in your diagnosis. In the background, you can also see evidence of fibroplasia with fibroblasts and collagen bands. MCTs can also degranulate, making it not as obvious on cytology, and in rare cases the granules may not stain, so keep that in mind if you suspect MCT. Work is being done to identify reliable characteristics to determine high-grade MCTs on cytology, including poor granulation as well as mitotic figures, multinucleation, nuclear pleomorphism, and anisokaryosis. However, MCT currently cannot be accurately graded on cytology, and this requires biopsy.

Plasmacytomas are usually well-circumscribed masses on the digits, ears, and mouths, generally of older dogs. The samples will generally be quite cellular, with discrete round cells that have variable amounts of basophilic cytoplasm. A key differentiating characteristic is that the cells will have a perinuclear clearing (the cell’s golgi zone), typical of plasma cells in general. Their nuclei are round or oval, eccentrically placed, and have indistinct nucleoli. Anisocytosis, anisokaryosis, and multinucleation are actually quite prominent, despite this typically being a benign tumor. There are a few unique things that you can see in a plasmacytoma that I think are especially cool. The first is flame cells, which have a pink, frilly corona (crown) around the outside of their cytoplasm. Next, there are mott cells, which have round basophilic Russell bodies in their cytoplasm, representing packets of globulin within that plasma cell. Finally, a small proportion of these tumors can have eosinophilic material within and around their cells, which is amyloid.

Histiocytomas are also quite common, usually in younger dogs, presenting as another type of well-circumscribed round mass on the skin, this one often being hairless and/or ulcerated. It occurs often on the head. The cells are round, lightly basophilic, with variably distinct borders. Nuclei are round, oval, or indented and have indistinct nucleoli. Anisocytosis and anisokaryosis are generally minimal. They often regress spontaneously, and it’s common to see small lymphocytes within these tumors that are in the process of regressing. I commonly struggle with differentiating plasma cell tumors and histiocytomas, but some differentiating features to look for include: perinuclear clearing (plasmacytoma), lymphocyte infiltration (histiocytoma), and occasional indented nuclei (histiocytoma). Histiocytomas also tend to have lighter cytoplasm, tend not to have many cells with more than two nuclei (plasmacytomas can), and anisokaryosis is rarely to a severe degree.

Histiocytic sarcomas originate from a different cell origin than histiocytomas, so thinking of them as the “malignant version of histiocytoma” probably isn’t quite right but will work for our purposes. They can be single or disseminated, with the localized ones being firm and often subcutaneous masses on the extremities. They are indeed malignant. They will often look very wonky on cytology. Their histiocytes have marked atypia, anisocytosis, and anisokaryosis, with multinucleated giant cells often seen. They often have plenty of cytoplasm and are often vacuolated. Sometimes they can also have a spindle shape. Their nuclei are round or indented and often have multiple nucleoli. HS is often the provider of some of the wildest cytologies you’ll see.

This makes up a decent starting list of some of the masses you can outwardly see, usually on the skin. There are obviously plenty more, and more things we can diagnose with cytology, but this can at least get us started.

References

Canine and Feline Cytopathology; Raskin, Meyer, Boes © 2023
eclinpath.com