chapter 9 FNA is used to evaluate palpable breast masses and cysts as well as nonpalpable mammographic abnormalities. FNA is highly accurate for palpable lesions,1,2 although its accuracy is limited with lesions smaller than 1 cm. Despite competition from the automated core needle biopsy (CNB) under stereotactic guidance, FNA delivers good results, especially in a multidisciplinary setting with on-site radiologists and pathologists.3–7 Complications of FNA are rare, the most common being bleeding. Occasionally, FNA causes partial infarction of the lesion, particularly fibroadenomas, which can hinder histologic confirmation of the diagnosis.8 For nonpalpable lesions, FNA is a useful technique for sampling cystic lesions with ultrasound guidance, whereas the CNB is more often used for mammographically identified calcifications.9 In pregnant or postpartum patients, FNA is preferred in order to avoid a draining, nonhealing wound that can result after a core or incisional biopsy. FNA is also useful for assessment of recurrent lesions. The accuracy of FNA of the breast, as with most things, is operator-dependent: Sensitivity for malignancy ranges from 65% to 98%, and specificity from 34% to 100%.1,10–16 False-positive results occur in 0 to 2% of cases.17 False-suspicious result rates are higher, ranging from 1% to 13%. In general, the sensitivity of FNA for palpable and nonpalpable malignant lesions (i.e., those sampled with mammographic or ultrasound guidance) is comparable.18–34 False-negative results occur because of errors in sampling, interpretation, or both.14,35 Some studies show that satisfactory specimens are more likely when pathologists rather than clinicians perform the aspiration.14,36–41 Whether clinician or pathologist, however, practice makes perfect, and the physician with more FNA experience obtains the more accurate result.42–44 The use of p63 and CK5/6 immunostaining increases the accuracy of FNA by helping distinguish well-differentiated carcinomas from benign lesions.45–48 A major advantage of FNA is the ease with which the sample can be assessed for adequacy.49,50 Although a touch imprint or wash of a CNB specimen can be done for rapid diagnosis, their utility is debatable.51–56 The use of FNA and/or CNB significantly decreases health care costs by decreasing the number of open surgical biopsies per breast cancer identified, without sacrificing early detection.57 When the diagnosis is benign, such as a lactating adenoma in a pregnant patient, FNA spares a patient with a solid and palpable lesion an open biopsy. A diagnosis of malignancy allows preoperative evaluation of available therapeutic options (lumpectomy with irradiation versus mastectomy), or it might persuade a reluctant patient to undergo surgical biopsy. FNA of the breast has its limitations. Although sensitive in detecting ductal carcinomas, it cannot distinguish between an in situ and an invasive ductal carcinoma. It cannot identify the presence of lymphatic or vascular invasion. It is less sensitive in tumors with low-grade cancer histology (e.g., tubular and lobular), papillary proliferations, and mucinous lesions.58–60 The diagnosis of lobular carcinoma and tubular carcinoma requires considerable experience in FNA interpretation25,59 even so, equivocal findings are common because of the benign cytologic appearance of such tumors.60 As with FNA of other sites, considerable discrepancy in performance exists among laboratories.61 In comparison with CNB, the nondiagnostic rate for FNA is higher, and FNA has a lower negative predictive value.55,58,62 Nevertheless, some authors have reported excellent results with breast FNA, and the combination of FNA and CNB may be superior to either alone.3,5,16,63–66 Although atypical ductal proliferative lesions are more amenable to classification by CNB than by FNA, false-negative diagnoses and underestimates of malignancy have been reported in a considerable percentage of cases even with CNB.67–70 Prognostic markers can be assessed with either technique.2,71 Many practitioners favor CNB over FNA62,72,73 because of the superior negative predictive value and wider acceptance of histopathology. The resulting increased use of CNB, particularly with stereotactic or ultrasound guidance, has led to diminished use of FNA,16,74 although some practitioners continue to use FNA for patients with radiologically and clinically negative lesions.75,76 The cost-effectiveness of FNA versus CNB is still debated.77–79 The increasing use of neoadjuvant chemotherapy in higher-stage cancers has further limited the use of FNA. Because an open biopsy is not performed, CNB, usually under ultrasonographic guidance, is superior because of its more robust sensitivity and ability to distinguish between ductal carcinoma in situ and invasive carcinoma. Furthermore, a CNB sample usually contains more abundant material for the determination of estrogen receptor (ER), progesterone receptor (PR), and HER2 status. Although a cell block prepared from an FNA sample permits multiple immunohistochemical stains, staining for PR in cell blocks may be discordant with tissue results, and staining for HER2 may be unreliable.80 Breast tumors are now classified in part on their molecular expression profile (luminal A, luminal B, HER2-positive, basal-like),81 and molecular-based risk stratification and prognostic testing by commercial products like Oncotype DX (Genomic Health) and MammaPrint (Agendia) is playing an increasing role in the management of patients.81 Entrance into many current neoadjuvant clinical trials stipulates core biopsy material for molecular testing. As new molecular tests are validated, the use of FNA may decline further unless studies incorporate FNA specimens in their design. FNA of axillary lymph nodes remains indispensable in the evaluation of lymph node status prior to neoadjuvant treatment or for preoperative staging,82–92 although some institutions are using CNB instead of FNA. A spontaneous nipple discharge not related to lactation or pregnancy is an abnormal finding. It may result from a breast lesion like a papilloma or a carcinoma or from a hormonal abnormality like that produced by a prolactin-secreting pituitary adenoma. Cytologic examination of a nipple discharge is generally used when the patient has no palpable or mammographic abnormality and is helpful in identifying small breast cancers and papillomatosis.93,94 If a palpable or mammographic abnormality is present, either FNA, CNB, or excisional biopsy is usually performed.95 The sensitivity of nipple discharge cytology ranges from 41% to 60%.96–98 Nipple discharge cytology is not useful as a screening test for breast cancer because a discharge can be obtained in only 7% to 14% of asymptomatic, nonpregnant, nonlactating women.96,98,99 In the future, biomarker analysis of nipple discharge fluid might help increase the sensitivity of cytology.100 A nipple discharge can be unilateral or bilateral; unilateral discharges are more likely to be malignant.96 The secretion can be milky, serous, purulent, or bloody. Cancer is most prevalent when the discharge is macroscopically bloody (4%) and less prevalent when it is purulent (0.8%), serous (0.2%), or milky (0.1%).96 Because bloody discharges are more likely than nonbloody discharges to contain malignant cells,94,101 some authors recommend cytologic examination of bloody secretions only, which represent 11% of all secretions.96 Others recommend examination of all discharges.98 Most patients with an intraductal papilloma do have a discharge,98 which may or may not be bloody. Although most patients with breast cancer do not have a discharge from the nipple, about 2% of breast cancers are detected by this method and by no other.98 Benign ductal cells are arranged in tight clusters that are small and spherical or large and branching; isolated benign ductal cells are very uncommon. Usually, the cells are small and have scant cytoplasm, but occasionally they are larger and have abundant cytoplasm. It is common for benign ductal cells to mold themselves around one another, giving the cluster a scalloped appearance. Foam cells are large histiocytes that are usually dispersed as isolated cells. They contain abundant vacuolated cytoplasm and a round or oval nucleus that is sometimes degenerated (Fig. 9.1A) When a secretion contains numerous groups of benign ductal cells, especially large, branching clusters, it is likely that the patient has an intraductal papilloma or a florid intraductal hyperplasia, lesions that can only be distinguished histologically. Figure 9.1 Nipple discharge cytology. To report FNA and nipple discharge results, the use of general categories with an implicit probability/risk of malignancy, followed by a specific diagnosis, is recommended.75,102–110 As might be expected, interobserver reproducibility is excellent for the positive category, poor for atypical, and fair to good for other categories.111 The exact wording for each category or diagnosis is less important than having a probabilistic (risk-based) system that clearly communicates the likelihood of malignancy. One pathologist might use the term “nondiagnostic” and another, “insufficient,” but the fact that this specimen is unsatisfactory is understood by clinician and pathologist alike. Nondiagnostic specimens contain too few well-preserved cells to permit an adequate evaluation—fewer than six epithelial cell clusters of at least 5 to 10 cells or less than 10 intact bipolar cells per 10 medium-power fields (×200).112,113 A negative (benign) diagnosis should be reserved for an adequate specimen with a minimum of five to six well-preserved benign ductal cell groupings. A negative FNA result is more reliable when a specific diagnosis corroborates a clinical and radiologic impression (e.g., fibroadenoma, lactating adenoma).102 The clinician should always correlate the cytologic result with the clinical findings and the mammographic impression (these constitute the so-called “triple test”) to reduce the risk of an undiagnosed malignancy, and clinical follow-up is indicated.113–115 The false-negative rate is greatly reduced when the triple-negative test is implemented.116 Needless to say, a negative cytologic result is by no means a guarantee that the nodule is benign, and a mammographically or clinically suspicious lesion necessitates a biopsy despite a negative cytologic result. The atypical category is used for a specimen with a low probability of malignancy. This category is unavoidable due to the significant overlap in the cytologic features of some benign and malignant entities. It generally requires biopsy assessment.117 The suspicious diagnosis is used for lesions that are probably malignant, but the atypical cells are too few, too poorly preserved, or too obscured by blood or inflammation for a definitive diagnosis, or when the findings suggest a type of breast cancer with minimal cytologic atypia, such as lobular carcinoma, tubular carcinoma,118 or papillary carcinoma. A histologic specimen should be obtained with any FNA sample that is deemed suspicious. Tightly cohesive three-dimensional aggregates: • fibroadenoma and phyllodes tumor • ductal proliferative lesions, from intraductal hyperplasia to ductal carcinoma in situ • well-differentiated ductal carcinomas Loosely cohesive three-dimensional clusters: Nuclear atypia is assessed on the basis of nuclear location, size, and shape; the chromatin pattern; and the quality of nucleoli. Although the standard cytologic criteria for malignancy (eccentrically placed, large, angulated, pleomorphic nuclei with irregular and large nucleoli) apply with moderately and poorly differentiated ductal carcinomas, some malignant tumors, including tubular, lobular, and mucinous carcinoma, show little nuclear atypia. The recognition of other features, like the architectural arrangement or the presence of abundant extracellular mucin, is important in the diagnosis of these tumors. The aspiration of breast cysts and the need for cytologic analysis is controversial. Aspiration collapses a cyst and is thereby therapeutic. The great majority of cyst fluids are benign; only about 2% prove to be carcinoma.119 Even complex cystic lesions are virtually always benign; in one study, only 0.3% were malignant.120 Furthermore, atypical cells can be seen in a cyst fluid, resulting in overdiagnosis and overtreatment when conservative follow-up would have been adequate.120,121 On the other hand, a small number of carcinomas are cystic and yield fluid that looks grossly much like that from benign cysts.122 If the fluid is not submitted for cytologic evaluation, these carcinomas will remain undiagnosed and untreated. It has been suggested that symptomatic complicated cysts, cystic lesions with thick indistinct walls and/or thick septations, intracystic masses, and predominantly solid masses with cystic degeneration are more likely to be malignant and thus merit cytologic or histopathologic examination.123 Nonproliferative lesions yield a scant specimen when the lesion is predominantly fibrous. When a cyst is present, the specimen is a fluid that may be thin and yellow or thick and darker in color. Apocrine cells line many but not all benign cysts. Apocrine cells have abundant granular cytoplasm that stains pink or green with the Papanicolaou stain and gray with a Romanowsky stain (Fig. 9.2). The nucleus is centrally located and round, with a prominent nucleolus, and moderate anisonucleosis is present in some cases. Benign apocrine cells are arranged as flat sheets; isolated cells are rare. Foam cells have abundant cytoplasm that is vacuolated rather than granular (see Fig. 9.1). Ductal epithelial cells are arranged in sheets and three-dimensional clusters (Fig. 9.3). Figure 9.2 Apocrine metaplasia. Figure 9.3 Benign ductal epithelium (nonproliferative fibrocystic changes). Granular cell tumors of the breast are rare. The cells of this tumor, thought to be of Schwann cell origin, have a very low nuclear-to-cytoplasmic ratio, a small nucleus, and a coarsely granular cytoplasm.124 The background is usually clean (Fig. 9.4). An apocrine carcinoma should be suspected when there is hypercellularity, marked nuclear atypia, and pronounced cell dyshesion, but a cautious diagnostic approach is advisable, because marked variation in nuclear size is also seen in apocrine metaplasia. Apocrine adenosis, although rare, can also manifest with nuclear atypia, but there is less nuclear hyperchromasia than with carcinoma, and there are many naked nuclei.125 Ductal proliferative lesions (i.e., proliferative FCC) comprise a group of lesions that vary in severity and degree of atypia. The spectrum includes proliferative lesions without atypia (“usual ductal hyperplasia”), atypical ductal hyperplasia, and atypical lobular hyperplasia. The criteria that define these entities and distinguish them from carcinoma in situ are histologic, not cytologic.126,127 Nevertheless, the degree of crowding and nuclear atypia allows for separation of the higher end of this spectrum from the lower. In one study, lesions reported cytologically as “proliferative lesion with atypia” were associated with a significantly higher frequency of histologically confirmed malignancywthan those reported as “proliferative lesion without atypia” (36.5% versus 1.7%).128 A lesion diagnosed as atypical by FNA should be considered for excision, because the morphologic features of well-differentiated invasive and in situ carcinomas overlap with those of benign entities.117 Figure 9.5 Ductal proliferative lesion without atypia. Figure 9.6 Ductal proliferative lesion with atypia. An intraductal papilloma is cytologically indistinguishable from proliferative FCC. Unlike proliferative FCC, however, many patients with an intraductal papilloma present with a nipple discharge or a discrete subareolar mass. Proliferative FCC may be impossible to distinguish from a fibroadenoma or phyllodes tumor,129 except that stromal fragments are fairly common in the latter two, and rarely seen in proliferative FCC. Pleomorphic carcinoma cells, calcium, necrosis, large nucleoli, and macrophages are indicative of comedo-type ductal carcinoma in situ and are usually diagnosed as either “suspicious” or “positive” (Fig. 9.7).130,131 Ductal carcinomas in situ of low and intermediate grade are usually interpreted as “atypical.”102 Some but not all well-differentiated ductal carcinomas in situ show more dyshesion than proliferative FCC. Still, distinguishing between ductal hyperplasia, atypical ductal hyperplasia, and well-differentiated ductal carcinoma in situ by cytology is difficult,19,104,117,131–135 even with the aid of image cytometry136,137 Although ductal carcinomas in situ are more likely than papillomas and other benign ductal lesions to have numerical chromosomal aberrations as detected by fluorescence in situ hybridization (FISH),138 it is unlikely that ductal proliferative lesions will be easily categorized by FNA in the near future. This limitation is not surprising, because primarily architectural, not nuclear or cellular, features define these lesions. The use of cell blocks may help in the cytologic classification of these entities.139 Figure 9.7 Suspicious for malignancy. Figure 9.8 Fibroadenoma, low magnification. • three-dimensional clusters with an antlerlike configuration (Fig. 9.9) Figure 9.9 Fibroadenoma. • bipolar cells and spindled/oval naked nuclei (Fig. 9.10) Figure 9.10 Fibroadenoma. Figure 9.11 Fibroadenoma. • some loss of epithelial cohesion • finely granular chromatin pattern Although no single criterion distinguishes a fibroadenoma from the ductal proliferations, a combination of features permits a distinction in most cases.140 In general, fibroadenomas are more cellular. Naked nuclei, although more abundant in fibroadenomas, are seen in both conditions. Stromal fragments and papillary antlerlike configurations, seen in many (but not all) fibroadenomas, are very uncommon in FCCs.140,141 The distinction between fibroadenoma and phyllodes tumor is difficult. Numerous individual, long, plump, spindle-shaped nuclei are characteristic of a phyllodes tumor.142,143 Also characteristic of phyllodes tumors are fibromyxoid stromal fragments with spindle-shaped nuclei and “fibroblastic pavements”: small fragments of cohesive fibroblasts forming a flat “pavement.”144 Hypercellular stromal fragments are more common in phyllodes tumor,143 but can be seen in fibroadenoma as well. The distinction between fibroadenoma and ductal carcinoma is usually straightforward; the most helpful diagnostic features are stromal fragments, antlerlike epithelial configurations, and honeycomb sheets of ductal cells, all of which are uncommon in ductal carcinomas. Some features can be misleading, however. Cytologic atypia is prominent in some fibroadenomas,140,145–147 and isolated cells with intact cytoplasm, a highly characteristic feature of ductal carcinoma, are seen in about 20% of fibroadenomas.8,140,147–149 Conversely, some ductal carcinomas masquerade as fibroadenomas. The greatest mimics are well-differentiated invasive ductal carcinoma and ductal carcinoma in situ. Naked nuclei, characteristic of fibroadenomas, are seen in some ductal carcinomas,148 although usually in fewer numbers than in a fibroadenoma. Nuclear hyperchromasia favors a diagnosis of malignancy, whereas nuclei with small, uniform nucleoli suggest fibroadenoma.148 The distinction is not discernible in all cases, and the differential diagnosis may be difficult, especially in older women.140,148,149 Another mimic of fibroadenoma is papillary carcinoma.150 Most of the isolated epithelial cells from carcinomas are round to oval with eccentrically placed nuclei, whereas those from fibroadenomas are elongated or columnar, with cytoplasm on both sides of the nucleus. Papillary carcinomas, however, can have isolated spindle-shaped or columnar epithelial cells on FNA. Smears with equivocal findings should be reported as atypical or suspicious. During pregnancy and lactation, the ductules of the terminal duct lobular unit become hyperplastic and manifest cytoplasmic vacuolization and luminal secretion. Occasionally, this change results in a discrete nodule, called a lactating adenoma, which is difficult to distinguish clinically from a malignancy. Because carcinoma is occasionally diagnosed in the setting of pregnancy, this diagnosis must be excluded in a pregnant or lactating woman. FNA in this setting may be especially useful, because a diagnosis of pregnancy-related or lactational changes could at least postpone and even spare the woman an excisional biopsy. • moderately cellular specimen • numerous isolated epithelial cells and/or stripped nuclei • nuclear enlargement without variation in size/shape • abundant delicate and wispy granular or finely vacuolated cytoplasm • cytoplasm easily stripped away, revealing: • occasional small ductal cell clusters and portions of lobules Figure 9.12 Pregnancy/lactational changes. The cells of invasive lobular carcinoma are similar in size to those of a lactating adenoma, but the foamy background, intact acini and lobules of benign breast tissue, and the prominent nucleoli of a lactating adenoma are absent in invasive lobular cancer. The nuclear size and shape of lactating adenoma cells can also resemble those of some well-differentiated ductal cancers, and the cytoplasmic features can overlap with secretory carcinoma.151 In general, ductal cancers do not have the foamy background characteristic of a lactating adenoma, and the cohesive groups of malignant cells in ductal cancers are not arranged in normal acinar structures. Also, the nuclei in ductal cancers are more hyperchromatic, the nucleoli less prominent, and the cytoplasm less wispy than in lactating adenoma. Non-Hodgkin lymphoma can resemble the changes of pregnancy and lactation. Both can have many isolated cells with prominent nucleoli, but the cytoplasmic features, proteinaceous background, and intact benign breast tissue typical of lactating adenomas are helpful. The isolated cells of lymphoma vary more in size and shape than those of a lactating adenoma.152 Fat necrosis can mimic carcinoma both clinically and mammographically and is commonly seen in patients who have had a previous surgical biopsy or other trauma to the breast. Fat necrosis is also encountered in male patients.153 Figure 9.13 Fat necrosis. The histiocytes seen in reactions to silicone injection or a ruptured silicone implant contain vacuoles that are larger than those seen in fat necrosis and often have a signet-ring appearance.154,155 In cryptococcosis of the breast, which can occur in an immunosuppressed patient, histiocytes have large cytoplasmic vacuoles containing refractile, budding yeast forms (Fig. 9.14). Figure 9.14 Cryptococcal infection. Some ductal carcinomas coexist with fat necrosis; they are identified by the presence of a distinct population of malignant cells in addition to histiocytes. The exceptionally rare lipid-rich carcinoma can also be confused with fat necrosis because the tumor cells have abundant vacuolated cytoplasm. Unlike fat necrosis, however, a lipid-rich carcinoma is hypercellular and shows marked nuclear atypia (Fig. 9.15). Figure 9.16 Radiation change. The histiocytic nuclei of fat necrosis are smaller than those of epithelial cells altered by radiation. Postsurgical and radiation-induced changes are a recognized pitfall but can usually be distinguished reliably from breast cancer.156–159 Most recurrent/persistent carcinomas demonstrate a moderately cellular or hypercellular specimen with many very atypical cells. Because radiation change also contains highly atypical epithelial cells, however, comparison with the morphology of the original tumor is recommended. Histologic confirmation may be necessary when the cytologic diagnosis is equivocal. The absence of isolated cells and necrotic cell debris is a useful finding, as these are more commonly seen in carcinomas. Acute mastitis usually is due to a bacterial infection and is seen most commonly in the postpartum period. Bacteria invade the breast through the small erosions in the nipple of a lactating woman, and formation of an abscess can result. Chronic mastitis can be a sequel to acute mastitis or, more commonly, associated with duct ectasia. Chronic mastitis is a disease of unknown etiology that results in the dilatation of large and intermediate-size ducts with a surrounding inflammatory infiltrate of lymphocytes and plasma cells. Some patients have a palpable mass that mimics carcinoma. A variant of chronic mastitis, characterized by an infiltrate composed predominantly of plasma cells, is called plasma cell mastitis. Granulomatous mastitis has the usual cytologic picture of granulomas and can be infectious (i.e., tubercular or fungal) in origin.160–162 The term granulomatous lobular mastitis has been given to a distinct clinical syndrome wyears after pregnancy. These lesions, which manifest as firm masses in the periphery of the breast, can be large and may suggest malignancy. The aspirate may contain cohesive clusters of histiocytes with “kidney bean” or boomerang-shaped nuclei. Other chronic inflammatory cells such as lymphocytes and plasma cells may be prominent. Special stains for bacteria, fungi, and acid-fast organisms are mandatory to rule out infection. Often called “recurring subareolar abscess,” this inflammatory condition arises in the areola as a result of squamous metaplasia of lactiferous ducts, with subsequent keratin plugging, dilatation, and rupture of the ducts. Without complete excision, the lesion can recur, potentially resulting in the formation of sinus tracts.
Breast
Specimen Types
Fine-Needle Aspiration
Nipple Discharge Cytology
A, Benign nipple discharge. Histiocytes (foam cells) have a kidney bean–shaped nucleus and abundant vacuolated cytoplasm (Papanicolaou stain). B, Suspicious nipple discharge. The sample contains very atypical cells with enlarged nuclei. The subsequent biopsy showed a high-grade comedocarcinoma (Papanicolaou stain).
Reporting Terminology
Evaluation of the Specimen
Lesions associated with architectural patterns
Benign Conditions
Cysts
Fibrocystic Changes
Nonproliferative Fibrocystic Changes
Large, flat sheets of apocrine cells have distinct cytoplasmic borders, a centrally located nucleus, and a prominent nucleolus. The abundant granular cytoplasm is gray-purple with a Romanowsky-type stain (A) and green with the Papanicolaou stain (B).
A tightly cohesive cluster of ductal epithelial cells without atypia is noted adjacent to apocrine metaplastic cells (Papanicolaou stain).
Proliferative Fibrocystic Changes
Note the interspersed myoepithelial cells, which stand out like sesame seeds on a bun (Papanicolaou stain).
In contrast with Figure 9.5, there is less regular nuclear spacing, more overlapping, and more prominent nuclear atypia, with a suggestion of cribriform spaces. Such proliferative lesions cannot be categorized precisely by FNA. Histologic examination revealed atypical ductal hyperplasia bordering on non comedo ductal carcinoma in situ (Papanicolaou stain).
The cells are loosely cohesive, with marked nuclear pleomorphism, prominent nucleoli, and a dirty background. Such specimens cannot be distinguished from invasive carcinoma by FNA. Histologic examination revealed comedo-type ductal carcinoma in situ (Papanicolaou stain).
Fibroadenoma
Cytomorphology of fibroadenoma
The specimen is hypercellular, with many folded sheets and antler-horn clusters (Papanicolaou stain).
A, Branching antler-horn clusters are the predominant arrangement of cells. There are rare stripped naked nuclei and bipolar cells in the background, but they are not prominent in this case, making it difficult to distinguish from a ductal proliferative process (Romanowsky stain). B, Clusters of tightly cohesive cells with minimal nuclear atypia are characteristic of fibroadenomas (Romanowsky stain).
Clusters of epithelial cells in a background of numerous stripped, elongated naked nuclei are characteristic of fibroadenomas. When stromal cells are absent, the diagnosis is more difficult (Papanicolaou stain).
Note the presence of nuclear atypia and prominent nucleoli. The tightness of the cluster is an important clue for avoiding an over-diagnosis of malignancy (Papanicolaou stain).
Pregnancy-Related and Lactational Changes
Cytomorphology of pregnancy and lactational changes (Fig. 9.12)
A, Numerous stripped (“naked”) nuclei are seen. B, Cells in loose clusters can also be seen. Nuclei are round or oval, with prominent nucleoli (ThinPrep, Papanicolaou stain).
Fat Necrosis
A, Histiocytes with abundant foamy (microvacuolated) cytoplasm are present (Romanowsky stain). B, An isolated histiocyte has abundant vacuolated cytoplasm. Smears of fat necrosis are typically sparsely cellular (Papanicolaou stain).
There are numerous intracellular yeast forms within macrophages (hematoxylin and eosin [H & E] stain).
Radiation Change
The cells show pronounced nuclear enlargement with concomitant cytomegaly. The nuclear-to-cytoplasmic ratio is thus maintained (ThinPrep, Papanicolaou stain).
Mastitis
Subareolar Abscess
Breast
Cytomorphology of benign nipple secretions
Cytomorphology of malignant nipple secretions (
General categories for reporting results of breast fine-needle aspiration
Low-power evaluation
High-power examination
Cytomorphology of nonproliferative fibrocystic changes
Differential diagnosis of nonproliferative fibrocystic changes
Cytomorphology of ductal proliferative lesion without atypia (
Cytomorphology of ductal proliferative lesion with atypia (
Differential diagnosis of ductal proliferative lesion without atypia
Differential diagnosis of fibroadenoma
Differential diagnosis of pregnancy and lactational changes
Cytomorphology of fat necrosis (
Differential diagnosis of fat necrosis
Cytomorphology of radiation change (
Differential diagnosis of radiation change
Cytomorphology
Cytomorphology of subareolar abscess
