Skin “Lumps and Bumps” Cytology

In small animal cytology, “lumps and bumps” refer to lesions that manifest within the cutaneous and subcutaneous layers of the skin, resulting in raised, conspicuous masses. These masses may be neoplastic or nonneoplastic and generally fall into 1 or more recognized categories^1-3:

• hyperplastic tissue
• inflammation/cellular infiltrates
• neoplasia
• cystic or fluid-filled masses

These lesions are ideal for cytologic sampling using fine needle aspiration (FNA), as they are accessed without advanced imaging (e.g., ultrasonography), and sedation or anesthesia is typically not necessary; they require only minimal equipment for sample acquisition; and they are often of high diagnostic yield.^1,3

Utility of Small Animal Cytology

Although the overall incidence of skin tumors in dogs and cats is challenging to determine, skin is a common site for neoplasia in small animal species. In the presence of a mass, cytology can often accurately distinguish neoplastic from nonneoplastic lesions, determine general categories of neoplasia, and lead to a more specific diagnosis.⁴ Multiple studies comparing cytologic with histologic diagnoses for neoplastic and nonneoplastic lesions have demonstrated the utility and reliability of cytology in veterinary medicine.^2,5

High concordance rates have been reported between cytology and histology in the evaluation of canine and feline cutaneous and subcutaneous lesions.^2,5 One retrospective study of 243 cytologic samples (198 from dogs and 45 from cats) agreed with histologic diagnoses on the presence or absence of neoplasia in 90.9% of cases when evaluated by clinical pathologists.² The same study reported a sensitivity of 89.3%, specificity of 97.9%, positive predictive value of 99.4%, and negative predictive value of 68.7% for the cytologic diagnosis of neoplasia.² An obvious advantage of achieving a cytologic diagnosis is avoiding additional diagnostics, thereby saving money for the client and enabling the prompt initiation of appropriate therapies.³ However, clients should be aware that biopsy and histologic evaluation may still be required for a more specific diagnosis. It is important to consider the experience of the evaluator and the influence of inflammation, which may mask or mimic neoplasia.⁴

Cytology, like all diagnostic tests, is not without limitations. The lack of tissue architecture restricts the amount of information cytology can provide, particularly when overlapping cytologic features (e.g., hyperplastic versus benign neoplastic lesions) exist. Acquiring a satisfactory sample also influences the usefulness of diagnostic cytology; inadequately cellular samples composed up to 16.8% of all cases in 1 study.² Therefore, appropriate sample acquisition, preparation technique, and screening for cellularity, along with resampling if needed, are recommended to avoid nonrepresentative or nondiagnostic samples.³ Detailed descriptions of how to obtain cytologic samples via FNA are provided elsewhere.^6,7

It is also important to remember that lesion type also affects the likelihood of a representative result.⁸ Knowledge of accurate and detailed patient history and mass characteristics significantly influences the cytologic interpretation and is critical (BOX 1).⁸ In many cases, histopathologic tissue evaluation should be pursued for a more definitive classification, with or without cytochemical, immunocytochemical, or immunohistochemical stains. Cytologic interpretation does not negate the need to submit a biopsy sample, as each yields essential information.

Approach to Sample Evaluation

The first step in evaluating a cytologic sample from a skin mass is assessing the entire slide at low-power magnification (4× to 10× objective) for adequate cellularity, distribution, cellular preservation, and staining quality (FIGURE 1).^3,9,11 Inadequately cellular samples may reflect poor sampling technique. Other considerations include cystic or poorly exfoliative lesions, particularly if resampling produces similar results. In such cases, a tissue biopsy may be required. If only cellular debris or necrotic tissue is identified, sampling a different site or histopathology of the mass is recommended.

Once sample quality has been confirmed, the next step is to evaluate all slides to determine the predominant cell population. Slides should be first assessed at low magnification to identify specific cell patterns or structures, which may provide information about the process or tissue type present.¹¹ High-power magnification (40×, 50×, or 100× objective) is then useful to scrutinize cellular morphology further, categorize the predominant cell type present (i.e., inflammatory or tissue cells), and identify criteria of malignancy or infectious agents.

Principal cell types determine classification, and evaluation for underlying causes (e.g., infectious or noninfectious etiologies) is warranted (TABLE 1). If only inflammatory cells are noted, inflammation would be suspected. An exclusively monomorphic tissue cell population with no significant inflammation suggests a hyperplastic or neoplastic population. Neoplasms are grouped into 4 general categories to assist in interpreting or generating appropriate differentials based on cellular arrangements and cytomorphology (TABLE 2).

Although the 4 basic categories of skin masses (i.e., hyperplasia, cystic, inflammation, and neoplasia) are based on the overall underlying process, it is essential to remember that they are not mutually exclusive. The same lesion may fall into more than 1 category (e.g., cystic adenoma, inflammatory carcinoma), and sampling multiple areas may reveal additional diagnostically significant information.

While cytology is most effective in differentiating neoplastic from nonneoplastic lesions, concurrent inflammation can induce dysplastic epithelial changes and/or reactive changes to fibrous tissue, mimicking neoplasia.^2,5,8 As such, extreme caution should be exercised in the presence of inflammation before diagnosing neoplasia, and consultation with a veterinary pathologist is advised. Overtly malignant neoplasms demonstrate specific structural changes that can be observed by light microscopy at high-power magnification (40× to 100× objective) as morphologic atypia, representing the basis of cytologic criteria of malignancy (TABLE 3).^9,10,15

Features of Common Cutaneous and Subcutaneous Lesions

Normal/Hyperplastic/Benign Neoplastic Tissue

Benign proliferative lesions (FIGURE 2A) are often cytologically and grossly indistinguishable, demonstrating absent to minimal cytologic criteria of malignancy. The same is true for certain malignant neoplasms, which can morphologically resemble a benign population (e.g., apocrine gland anal sac adenocarcinoma) despite an aggressive biologic behavior (FIGURE 2B). In these cases, histopathologic evaluation is required for differentiation when neoplasia is suspected clinically and further characterization is desired.

Figure 2A. Sebaceous epithelium. Mature sebocytes (orange arrow) surrounded by cuboidal, basilar epithelium (black arrow). Normal tissue, hyperplasia, and benign neoplasia (sebaceous adenoma) appear identical cytologically and require histopathologic evaluation for further characterization. Wright-Giemsa stain, 40× objective.

Figure 2B. Apocrine gland anal sac adenocarcinoma. Cytologic samples are typically highly cellular and composed of sheets of loosely cohesive cuboidal epithelial cells with indistinct cell borders, occasionally forming trabeculae or rosette-like structures. Often, cells are ruptured and appear as bare/naked nuclei (neuroendocrine appearance) within a pool of shared basophilic cytoplasm. Intact cells contain scant amounts of cytoplasm and central round or slightly oval nuclei with coarsely clumped chromatin and 1 to 3 small nucleoli. Anisocytosis and anisokaryosis are modest despite the aggressive biologic behavior of this tumor. Wright-Giemsa stain, 50× objective.

Cystic or Fluid-Filled Lesions

Cystic or fluid-filled skin lesions are often tumor-like lesions and may be associated with tissue injury from trauma (e.g., seroma, hygroma, hematoma, necrosis, abscess). Cytologic samples typically contain few nucleated cells (except for abscess) due to an absence of lining tissue cell exfoliation, and location can be beneficial for further differentiation. Seromas, for example, are often found at sites of previous tissue disturbance (e.g., prior surgery sites), hygromas are associated with bony prominences/pressure points, and follicular or glandular cysts are located within the skin. Follicular cysts demonstrate characteristic cytologic features such as cholesterol crystals, anucleate squamous epithelial cells, and keratin. Similar features are found in cystic follicular or keratin-producing tumors (e.g., pilomatricoma, trichoblastoma); therefore, distinction using cytology alone is not always possible. Necrosis and fibroplasia are also seen in various lesions as a response to tissue injury, inflammation, or neoplasia.

Not surprisingly, hematomas (FIGURE 2C) contain an abundance of erythrocytes, as well as blood-derived leukocytes and vacuolated macrophages, some demonstrating erythrophagia. Heme-breakdown products, hemosiderin, and hematoidin appear within 12 to 24 hours, suggesting chronicity. Notably, some neoplasms have large blood-filled cavities, and neoplastic cells may not exfoliate (e.g., hemangioma, hemangiosarcoma, telangiectatic osteosarcoma); thus, the absence of overtly neoplastic cells does not exclude the possibility of neoplasia.

Figure 2C. Canine hematoma. Vacuolated macrophages contain phagocytized red blood cells (arrow) and/or dark blue to black red blood cell degradation products (hemosiderin; asterisk). Hematoidin crystals (not shown) may also be seen and appear as rhomboid, yellow/orange crystalline structures in tissues with low oxygen tension. Wright-Giemsa stain, 50× objective.

Inflammation

As mentioned, inflammation may be due to infectious or noninfectious processes, and determining the predominant cell type aids in generating diagnostic differentials. Although infectious etiologic agents are not always identified, cellular composition (e.g., predominately neutrophilic versus neutrophilic and macrophagic) and characteristics such as epithelioid macrophages (FIGURE 3A) and multinucleated giant cells (FIGURE 3B) can provide clues to the type of agents that might be present, helping to select the next appropriate step (e.g., bacterial versus fungal culture). In blastomycosis (FIGURE 3C), for example, fungal yeast elicits a mixed population of neutrophils, some demonstrating degenerate features (karyolytic, karyorrhectic, or pyknotic nuclei), and macrophages (including epithelioid macrophages and multinucleated giant cells), representing pyogranulomatous inflammation. Thus, seeing this type of inflammation should alert slide evaluators to consider the possibility of a fungal or bacterial agent (Figure 3D), diligently look for infectious organisms, and, if not found, pursue additional testing (e.g., repeated aspiration, biopsy with histology, fungal culture, serology).

Figure 3A. Pyogranulomatous inflammation. Pyogranulomatous is reserved for inflammatory populations composed of neutrophils and epithelioid macrophages (circle) ± multinucleated giant cells. Wright-Giemsa stain, 50× objective.

Figure 3B. Pyogranulomatous inflammation. Multinucleated giant cells seen in a lesion with epithelioid macrophages, degenerating macrophages, and fungal yeast (not shown). Wright-Giemsa stain, 50× objective.

Figure 3C. Blastomycosis. Degenerate neutrophils surround a broad-based budding yeast, compatible with Blastomyces dermatitidis yeast. Wright-Giemsa stain, 100× objective.

Figure 3D. Septic suppurative inflammation. Degenerate neutrophils contain intracellular bacteria. Wright-Giemsa stain, 100× objective.

Neoplasia

Neoplastic populations are categorized based on cytomorphology (TABLE 2) and include (1) round/discrete cells, (2) epithelial cells, (3) mesenchymal cells, and (4) bare/naked nuclei (FIGURE 4). Melanocytic neoplasms, which can demonstrate variable morphology and be challenging to classify cytologically, are a notable exception.¹⁶

Figure 4A. Round cell tumor (canine plasmacytoma). Often highly cellular, composed of individualized round cells with an absence of cell junctions or extracellular matrices. In this case, well-differentiated plasma cells are characterized by round, eccentrically placed nuclei with fine to coarse chromatin, inconspicuous nucleoli, and abundant blue cytoplasm with a distinct perinuclear clearing (orange arrow) identifying the Golgi apparatus. Rare intracytoplasmic immunoglobulins (Russell bodies; asterisk) and multinucleation may also be present (square). Wright-Giemsa stain, 100× objective.

Figure 4B. Epithelial neoplasia (carcinoma). Cells are cohesive due to tight junctions/desmosomes (orange arrow), forming sheets or 3-dimensional clusters. Wright-Giemsa stain, 100× objective.

Figure 4C. Mesenchymal cell neoplasia. Mesenchymal cell neoplasms (soft tissue sarcoma) are typically poorly cellular and exfoliate as individual cells or in variably sized aggregates, often associated with the matrix (not shown). Wright-Giemsa stain, 50× objective.

Figure 4D. Bare/naked nuclei neoplasm (neuroendocrine, thyroid carcinoma). These types of tumors are composed of loosely adherent cells and bare nuclei attributed to the fragile nature of the cells of origin (e.g., endocrine, neural, neuroendocrine). Wright-Giemsa stain, 20× objective.

Cells should also be evaluated for cellular atypia (i.e., general and nuclear criteria of malignancy; TABLE 3, FIGURE 5). However, malignant cells cannot be distinguished from benign cells based on a single feature. Only when the majority of cells demonstrate numerous criteria (typically at least 3 to 5 nuclear criteria) in an adequately cellular and preserved sample should malignancy be suspected. Nuclear criteria are more significant than general criteria and unlikely to be influenced by external factors such as inflammatory mediators. Biopsy and histopathologic evaluation should be considered if ambiguity or concurrent inflammation exists.

Figure 5A. Multinucleation and prominent, variably sized and shaped nucleoli (circles). This lesion was identified as canine hemangiosarcoma. Wright-Giemsa stain, 100× objective.

Figure 5B. Multinucleation, intracellular anisokaryosis, high nuclear to cytoplasmic ratio, and nuclear molding (arrow). This lesion was identified as carcinoma. Wright-Giemsa stain, 100× objective.

Figure 5C. Bizarre/abnormal mitotic figure from a canine prostate aspirate. Chromosomes are improperly aligned. This sample showed numerous criteria of malignancy and neutrophilic inflammation. The nuclear criteria of malignancy, however, were more consistent with carcinoma. Wright-Giemsa stain, 100× objective.

Criteria of malignancy can aid in predicting the biological behavior of a neoplasm (benign or malignant). It is, however, critical to recognize that morphology is not always indicative of biological behavior, and dysplastic features induced by inflammation/ulceration/necrosis can appear similar. Alternatively, as previously mentioned, a benign cytologic appearance does not always indicate a concurrent benign biologic behavior. In such cases, histopathology or repeat aspiration may be necessary if the mass persists following the resolution of inflammation or if there are other indicators that this neoplasm has a more aggressive clinical course (e.g., nodal or distant organ involvement, known malignant behavior despite morphology).

Summary

Cytology is a tool that can be used to determine the underlying pathology of skin masses, resulting in a definitive diagnosis or an appropriate list of differentials. It is not a replacement for biopsy and histopathology; rather, these approaches should be used together, with or without cytochemical, immunocytochemical, or immunohistochemical stains. By understanding the limitations of cytology, the importance of sample quality, and the role of cytology in the workup of patients, veterinarians can use the knowledge gained through cytologic assessment to inform clinical decision-making regarding further diagnostics or treatment considerations for these lesions.