Recurrent challenges in the evaluation of fibroepithelial lesions.
Abstract: Context. -- The morphologic spectrum of mammary fibroepithelial lesions ranges from fibroadenoma, a common benign neoplasm, to phyllodes tumor, an uncommon lesion that can sometimes recur and metastasize.

Objective. -- To focus on problems encountered in the diagnostic evaluation of fibroepithelial tumors, highlighting the diagnostically relevant morphologic features and providing an update on the immunohistochemical profile and genetic alterations of these rare neoplasms.

Data Sources. -- A PubMed search of the English-language literature identified published reports on fibroepithelial lesions, with a special focus on phyllodes tumor. The results and conclusions of these studies form the basis of this review.

Conclusions. -- The distinction between fibroadenoma and phyllodes tumor is usually not problematic, especially in excision specimens. In some cases, however, the diagnostic evaluation of fibroepithelial lesions can be challenging, especially if only limited material is available. Morphologic predictors of local recurrence of phyllodes tumor include cellularity and cytologic atypia, mitotic activity, positive margins, infiltrative borders, fibroproliferative satellite nodules, and past history of fibroadenoma. Predictors of distant metastasis include size, tumor necrosis, and stromal overgrowth. None of these parameters, however, constitutes a definite marker of malignancy. Presently, molecular and immunohistochemical techniques play a limited role in the diagnosis of fibroepithelial lesions.
Article Type: Report
Subject: Breast cancer (Diagnosis)
Breast cancer (Development and progression)
Epithelial tumors (Diagnosis)
Epithelial tumors (Development and progression)
Biopsy, Needle (Usage)
Metastasis (Evaluation)
Immunohistochemistry (Analysis)
Author: Giri, Dilip
Pub Date: 05/01/2009
Publication: Name: Archives of Pathology & Laboratory Medicine Publisher: College of American Pathologists Audience: Academic; Professional Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2009 College of American Pathologists ISSN: 1543-2165
Issue: Date: May, 2009 Source Volume: 133 Source Issue: 5
Topic: Canadian Subject Form: Epithelial tumours; Epithelial tumours
Accession Number: 230151992
Full Text: A mammary fibroepithelial lesion is a neoplastic proliferation of the specialized stroma of the breast and secondarily distorts lobules and ducts, incorporating them within the mass. The resulting tumor contains epithelial elements, but only the stromal component is neoplastic. Fibroepithelial lesions encompass a wide variety of benign and malignant lesions ranging from fibroadenoma (FA) to malignant phyllodes tumor (PT).

FIBROADENOMA

Fibroadenoma is the most common benign breast neoplasm. A typical FA (Figure 1) is a well-circumscribed mass characterized by regular and symmetric distribution of epithelial and fibrous components. In an FA, the stroma shows low cellularity and lacks mitotic activity, and the spindle cells are cytologically bland. Fibroadenoma grows slowly and rarely reaches a large size. Two unusual variants of FA are worth special mention. In myxoid FA, the stromal component is hypocellular and shows extensive myxoid changes, with bland, spindled to stellate cells. Myxoid FA has been reported in the context of Carney syndrome, (1) but it can also occur sporadically, and it remains unknown how often mammary myxoid FA constitutes the herald of this disease. Juvenile FA, on the other hand, shows higher stromal cellularity and prominent epithelial hyperplasia, features uncommon in a typical FA. Mitotic activity has also been described in juvenile FA. (2) Nonetheless, the circumscribed border, lack of cytologic atypia, stromal overgrowth, and necrosis support its benign nature. Juvenile FA can assume a large size, and the terms cellular fibroadenoma or giant fibroadenoma have been used by some authors for this entity. (3)

PHYLLODES TUMOR

Phyllodes tumor was originally described in 1838 by Muller, (4)f who believed the lesion to be benign but called it cystosarcoma because of its cystic change and fleshy cut surface. As many as 62 different synonyms have been used for PT during the past 150 years, (5) a historic testament to the surgical pathologist's inability to capture the true nature of this lesion. The term PT was introduced for the first time in 1960 by Lemonaco. (6) The first example of a malignant PT (or malignant cystosarcoma, as it was then called) was described in 1931 by Lee and Pack (7), nearly 100 years after Muller's original report. Foote and Stewart (8) were the first to recognize that the malignant changes of PT involve the connective tissue elements and not the associated epithelial component. In 1951, Treves and Sunderland (9) separated PT into the 3 subclasses of benign, borderline, and malignant, and this subclassification continues to be followed today, integrating multiple pathologic features (Table). (3) Nonetheless, accurate classification of PT continues to be challenging on many occasions, partly because PTs are morphologically heterogeneous, with different areas of the same tumor showing different appearance, including areas morphologically indistinguishable from FA.

[FIGURE 1 OMITTED]

Phyllodes tumor is characterized by marked expansion of the specialized mammary stroma, with consequent relative reduction of the epithelial component (Figure 2, A). The stroma grows into the ducts (so-called intracanalicular growth pattern), with formation of leaflike projections, pathognomonic of this tumor (Figures 2, B and 3), and hence the term phyllodes (Greek: phyllos = leaf) used to describe this lesion. Phyllodes tumor typically shows infiltration into the adjacent tissue, best appreciated at the tumor periphery (Figure 4), and its extent is an important parameter in the subclassification of PT. Other diagnostic parameters include mitotic rate, cellularity, and nuclear atypia. No definitive consensus exists on the number of mitoses required for classification into each of the 3 subgroups of PT. For example, the World Health Organization classification of [PT.sup.3] (Table), provides no numeric cutoff to distinguish benign and borderline PT and requires more than 10 mitoses per 10 high-power fields (HPF) for malignant PT. On the other hand, some authors (2) require fewer than 2 mitoses per 10 HPF for benign PT, 2 to 5 per 10 HPF for borderline PT, and more than 5 per 10 HPF for malignant PT. Stromal overgrowth, defined as the absence of any epithelial component in at least one X40 field (Figure 5), is commonly seen in malignant PT, and it can rarely also occur in borderline PT.

FINE-NEEDLE ASPIRATION CYTOLOGY AND NEEDLE CORE BIOPSY IN THE PREOPERATIVE DIAGNOSIS OF

FIBROEPITHELIAL LESIONS

The question "Fibroadenoma or Phyllodes tumor?" constitutes a recurring dilemma in the clinical practice of every pathologist dealing with breast biopsies. In these cases, the diagnosis will greatly influence clinical management of the lesion ranging from observation alone, an acceptable option for FA, to excision with wide negative margins for PT.

The pathologist's first encounter with a fibroepithelial lesion typically involves review of limited material, either a core biopsy or, more rarely, a fine-needle aspiration (FNA) cytology specimen.

In a recent review, Jacklin et al (10) examined the role of FNA in the diagnosis of PT. The authors reviewed the findings in 9 studies (11-19) with a total of 201 patients. The overall rate of definitive diagnosis was 63%, (range, 33%-77%). Cellular stromal fragments were highly suggestive of PT, as well as dispersed stromal cells with elongated and spindled nuclei constituting more than 30% of the sample. Large (>1 mm), elongated, and wavy epithelial clusters were also a feature of PT, in contrast to smaller tubular or blunt-branching clusters found in FA. Even though the authors identified these cytomorphologic features as useful in the diagnosis of PT, they concluded that FNA has a high false-negative rate, resulting in underdiagnosis of at least 25% of PT. Hence, even if the cytologic diagnosis of PT is possible in a proportion of cases, many PTs will be misdiagnosed as FAs if FNA is used as the only diagnostic modality. The low accuracy of FNA-based diagnosis is secondary to lack of architectural detail and limited sampling, given that PTs are characterized by heterogeneous distribution of stromal cellularity, cytologic atypia, and secondary epithelial component.

Core needle biopsy is currently the preferred diagnostic modality in the initial evaluation of breast lesions, but only a few studies have critically assessed its role in the diagnosis of fibroepithelial lesions. (20-25) In a study by Dillon et al, (20) 9 of 23 patients diagnosed with PT on excision had a prior core biopsy diagnosis of FA or "benign," amounting to a false-negative rate of 39%. In a study of 57 patients with fibroepithelial lesions diagnosed on core biopsy, (21) the diagnosis of FA was favored in 25 patients and confirmed on excision in 23, whereas 2 patients were found to have PT. Similarly, 19 of 23 patients with core biopsy diagnosis favoring PT had PT on excision, but 4 had FA. In this study, 5 of 9 patients with equivocal findings on core biopsy had FA and 4 had PT on excision. Nonetheless, the authors concluded that core biopsy remains useful in the evaluation of fibroepithelial lesions and contributes to reduce the need for re-excision. No specific morphologic criteria for the diagnosis of FA or PT on core biopsy were defined in this study.

[FIGURE 2 OMITTED]

Jacobs et al (22) assessed the findings in the excision specimens of 29 fibroepithelial lesions showing cellular stroma on core biopsy. Four tumors showing mild cellularity on core biopsy were FA on excision, whereas all those with marked stromal cellularity were PT. Lesions with moderately cellular stroma on core biopsy were FA in 12 cases and PT in 8. The finding of stromal mitoses was the only morphologic discriminator between the 2 types of tumor. Immunoreactivity for Ki-67 was also significantly higher in PT than in FA (6% vs 1.6%, respectively).

Lee et al (23) systematically evaluated morphologic parameters in core biopsies of 44 patients diagnosed with PT on excision and 38 patients with FA. Mitotic activity was absent in FAs but detectable in nearly one-third of PTs. Fragments of stroma lined by epithelium on one or two opposing edges (Figure 6) were present in almost one-third of core biopsies of PTs, but not in those of FAs. Stromal overgrowth or stromal expansion (defined as absence of epithelium in a X100 field) was noted in only 1 of 38 FAs but occurred in 36% of PTs. Moderate to marked stromal cellularity was evident in nearly a third of core biopsies of PTs but not in those of FAs. Similarly, infiltrative edges were not seen in core biopsies of FAs but were noted in nearly half of the biopsies of PTs. Another significant finding was the presence of entrapped fat, noted in one-third of the PT cases but not in FAs.

Yohe and Yeh (24) compared the findings in core biopsies of 15 PTs with those of 16 FAs and identified high Ki-67 labeling index (4.8% vs 0.6%), increased mitotic rate, tissue fragmentation, pronounced stromal cellularity, especially around glands, and stromal overgrowth as features suggestive of PT. Extensive sampling of a large tumor was also more likely to yield a correct diagnosis. Foxcroft et al (25) specifically commented on difficulties in obtaining accurate preoperative diagnosis of fibroepithelial lesions. In their study of 84 cases of PT (confirmed on excision specimens), this differential diagnosis was raised preoperatively in only 23% of cases on FNA and in 65% on core biopsy.

Review of the literature consistently underlines that accurate distinction between FA and PT may not be possible in all instances when only limited material is evaluated. Morphologic features favoring PT over FA include mitotic activity, increased cellularity, presence of stromal overgrowth, infiltrative borders, tissue fragmentation (Figure 6), and higher Ki-67 proliferation index. Large size (>3 cm) and rapid tumor growth also suggest a serious lesion, even when the morphologic features in the core biopsy are those of an FA. In such situations, it is appropriate to excise the lesion for complete evaluation.

Use of the above criteria in the prospective diagnosis of core biopsy material likely will improve diagnostic accuracy, but no such study is available at present.

The following example illustrates the need for a pragmatic approach in the diagnosis of fibroepithelial lesion in the setting of a needle biopsy. Figure 7, A, shows a cellular fibroepithelial lesion in a core biopsy. The stromal cells do not show mitotic activity but have cytologic atypia. A diagnosis of cellular FA was rendered based on the lack of stromal overgrowth or mitotic activity. The lesion grew rapidly in the ensuing 2 months and had to be excised with a portion of skin. As shown in Figure 7, B through D, this lesion revealed marked stromal overgrowth and abundant mitotic activity (>10 per 10 HPF). There was also evidence of tumor necrosis. It is obvious that although the biopsy was interpreted as a cellular FA, the lesion, in fact, was a malignant PT. Two years later, the patient developed lung metastases (Figure 7, E and F). This case highlights the need for proper clinicoradiologic correlation while assessing fibroepithelial lesions in core biopsy specimen and the need to avoid misleading terms, such as "cellular" FA, especially when only limited material is available.

MORPHOLOGIC PREDICTORS OF LOCAL RECURRENCE AND SYSTEMIC METASTASIS

Although PTs are classified into the 3 categories of benign, borderline, and malignant, it remains very difficult to predict the behavior of these lesions using morphologic features alone. The rate of local recurrence of PT ranges from less than 20% for benign PTs to more than 25% for malignant PTs. Distant metastasis occur in less than 5% of borderline PTs to about 25% for malignant PTs. (2) Regional lymph node metastases are extremely rare, and axillary dissection is not part of the surgical management of PT. (26) Lester and Stout (27) were the first to note that biopsy appearances alone may not provide the basis for accurate classification of the "biological potential" of PTs. Oberman (28) also emphasized that PT shows variable histology and has an unpredictable clinical behavior, limiting the accuracy of histologic assessment of malignancy. Norris and Taylor (29) studied the morphologic features of PT from 94 patients and concluded that tumors smaller than 4 cm and with fewer than 3 mitoses per 10 HPF were unlikely to cause a patient's death. They also suggested that a pushing margin and minimal cytologic atypia were associated with very low risk of local recurrence and/or lethal outcome. These authors emphasized that no single morphologic feature could provide a clear-cut separation of benign and malignant PT. Moffat et al (30) studied features associated with local recurrence in a group of 32 PTs and showed that the presence of tumor at margins was the major determinant, whereas other histologic features did not show similar relevance.

[FIGURES 3-6 OMITTED]

de Roos et al (31) studied 38 patients with PT and reported that inadequate preoperative diagnosis based on mammography, cytology, and core biopsy frequently led to local excision with positive margins. Local recurrences (one or more times) occurred in 9 patients who did not undergo re-excision to obtain negative margins. The risk of local recurrence was regardless of the tumor histologic grade. Kapiris et al (32) confirmed that tumor size greater than 10 cm and positive margin status were significant predictors of local recurrence and distant metastases in 48 patients with morphologically malignant PT. Hawkins et al (33) identified stromal overgrowth as the most important predictor of distant metastasis in a series of 33 patients with PT. Mitotic activity (>10 per 10 HPF), nuclear pleomorphism, and infiltrating margins were also important parameters, and local recurrence predisposed to the development of distant metastasis.

[FIGURE 7 OMITTED]

[FIGURE 8 OMITTED]

Barrio et al (34) recently reported the findings in a series of 293 PTs with long-term follow-up. Local recurrence rate was significantly increased in patients with positive margins versus those with negative margins (18% vs 10%, respectively). Tumor necrosis and "fibroproliferative" foci in the breast tissue adjacent to the index lesion were associated with increased risk of local recurrence. In this study, fibroproliferation was defined as the presence of fibroadenomatous "satellite" nodules and/or periductal stromal proliferation adjacent to the main lesion (Figure 8). The study also noted a trend toward higher local recurrence in patients with a prior history of ipsilateral FA, although the number of cases was too low to reach statistical significance. No difference was found in the local recurrence rate of malignant and nonmalignant forms of PT, confirming prior reports. (30, 31) Death due to PT was a rare event (2%) and occurred only in patients whose tumors showed more than one worrisome pathologic feature, including large tumor size (>7 cm), infiltrative borders, marked stromal overgrowth, marked stromal cellularity, high mitotic count, and tumor necrosis. The latter parameter was identified as a specific independent risk factor for distant metastasis for the first time in this study. The low mortality rate of PT reported by Barrio et al (34) is not surprising, and Grabowski et al (35) already had noted that, overall, malignant PT has a less aggressive course than invasive breast carcinoma.

IMMUNOHISTOCHEMISTRY AND MOLECULAR STUDIES IN THE EVALUATION OF PT

Several Immunohistochemical Markers Have Been Used in the Evaluation of PT

Kleer et al (36) assessed Ki-67 (MIB-1) and p53 staining in 20 PTs and concluded that neither marker could reliably predict local recurrence, although Ki-67 was helpful in distinguishing benign from malignant PT in diagnostically difficult cases. Feakins et al (37) also found no significance of p53 immunoreactivity with regard to tumor recurrence or long-term survival in a series of 57 PTs. On the other hand, Yonemori et al (38) have recently reported that p53 expression and MIB-1 index may be significant prognostic factors in patients with PT. Based on these reports, it appears that at present there is no consensus on the usefulness of p53 and Ki-67 as risk predictors of PT.

Sawyer et al (39) studied c-kit immunoreactivity in PT and found strong c-kit expression in 4 of 8 cases of malignant PT, but lack thereof in 20 benign PTs. Tan et al (40) showed immunohistochemical positivity for c-kit in 7 of 31 locally recurrent PTs (22.5%), in contrast to 10 of 242 PTs (4%) that had not recurred. These studies suggest that although c-kit is expressed in a small proportion of PTs, this molecule may be important in predicting local recurrence, but further confirmation is required.

Tse et al (41) evaluated CD10 in fibroepithelial lesions and found it rarely in FA (1/33) and benign PT (6/102), but more frequently in borderline PT (16/51) and malignant PT (14/28). Despite these interesting results, to the best of our knowledge, none of these markers finds application in the routine diagnostic work-up of fibroepithelial lesions.

The use of immunoperoxidase studies is, however, very important to avoid possible pitfalls in the diagnosis of PT. Occasionally, a spindle cell carcinoma can focally entrap normal epithelium in a pattern that mimics a fibroepithelial lesion, as recently reviewed by Lee. (42) The use of immunoperoxidase markers, such as keratins, myoepithelial markers, CD34, and [beta]-catenin, can be helpful in the differential diagnosis of mammary spindle cell lesions, including spindle cell carcinoma, fibromatosis, and PT. Scattered focal expression of keratins can occasionally be seen in PT, as described by Lerwill (43); hence it is important to always use a wide panel of antibodies to accurately characterize a morphologically indistinct spindle cell lesion. In particular, focal distention of residual ducts can occur in a spindle cell carcinoma and closely mimic a PT; and vice versa, the epithelial component associated with a PT can be minimal, and occasionally difficult to appreciate. Figure 9 illustrates an example of a spindle cell lesion with fibromatosis-like areas and entrapped mammary ducts. Although a malignant PT would be in the differential diagnosis, the spindle cells express basal keratins (cytokeratins 5/6 and 14) and the myoepithelial marker p63, supporting the diagnosis of a spindle cell metaplastic carcinoma rather than PT.

[FIGURE 9 OMITTED]

Layfield et al (44) used DNA flow cytometry to assess tumor ploidy, but they could not establish any relationship with tumor histology, size, local recurrence, and distant metastasis. Lae et al (45) examined the chromosomal alterations in fibroepithelial tumors using comparative genomic hybridization and found 2 distinct patterns of genomic imbalance in PT. The benign pattern was characterized by few or no chromosomal changes, whereas the malignant pattern displayed numerous and recurrent chromosomal alterations, including gain of 1q and loss of 13q. Jones et al (46) have provided a major advance in our understanding of the genetic profile of PT using array comparative genomic hybridization-based technology. These authors found that large-scale genetic changes seen in malignant PT, such as loss of [p16.sup.INK4A], are not present in benign PT. Some malignant PTs also showed mutation and methylation of [p16.sup.INK4A], resulting in defective function of this molecule. These findings suggest an important role for [p16.sup.INK4A] in the biology of malignant PT. In the same study, recurrent tumors showed genetic changes not present in the primary lesion. In particular, recurrent "benign" tumors seemed to acquire the molecular phenotype of borderline/malignant lesions. This finding is consistent with the observation that histologic progression occurs in PT, as seen in recurrent and/or metastatic PT (Figure 10).

SUMMARY

Phyllodes tumors are uncommon mammary neoplasms, and their evaluation can sometimes constitute a diagnostic challenge. Phyllodes tumor can focally resemble FA, leading to diagnostic underinterpretation, particularly on evaluation of limited material, such as FNA or needle core biopsy. As a pragmatic approach, it is prudent to recommend excision of any fibroepithelial lesion that shows one or more morphologic findings typical of PT, such as increased stromal cellularity, infiltrative borders, mitotic activity, stromal overgrowth, necrosis, cytologic atypia, and tissue fragmentation, in needle core biopsy material. Excision is also indicated if the fibroepithelial lesion is large (>3 cm in size) or is rapidly growing. The proliferation marker Ki-67 can be a helpful aid in the distinction between FA and PT, with the caveat that its expression tends to be low in benign and borderline PTs, and thus it does not constitute an absolute discriminator. Excision with widely negative margin is important to prevent local recurrence, regardless of the histologic grade of PT. Large tumor size, stromal overgrowth, mitotic activity, necrosis, a high degree of cytologic atypia, and the presence of foci of "fibroproliferation" indicate a higher risk for recurrence and or metastases. A trend toward increased recurrence risk if the patient has a prior history of FA has also been reported.

Overall, the risk of death due to malignant PT is quite low (as low as 2%), and morphologic criteria are insufficient to predict tumor behavior.

Use of immunohistochemical markers, such as CD10 and c-kit, may provide useful information, but more comprehensive studies involving a large number of cases and long-term follow-up are required before these markers gain an established place in the work-up of PT. Immunohistochemistry plays a role also in ruling out spindle cell carcinomas, which can occasionally masquerade as a PT.

[FIGURE 10 OMITTED]

Genetic studies suggest that the benign and malignant PTs have distinct molecular profiles and also support evidence of progression in metastatic lesions. Future studies focused on the genetic profile of PT could lead to the identification of novel therapeutic targets.

Accepted for publication January 20, 2009.

References

(1.) Carney JA, Toorkey BC. Myxoid fibroadenoma and allied conditions (myxomatosis) of the breast: a heritable disorder with special associations including cardiac and cutaneous myxomas. Am J Surg Pathol. 1991;15:713-721.

(2.) Rosen PP. Fibroepithelial neoplasms. In: Rosen PP, ed. Rosen's Breast Pathology. 2nd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2001:163-200.

(3.) Bellocq JP, Magro G. Fibroepithelial tumors. In: Tavassoli FA, Devilee P, eds. World Health Organization Classification of Tumors: Tumors of the Breast and Female Genital Organs. Lyon, France: IARC Press; 2003:99-103.

(4.) Muller J. Uber den feinern Bau und die Formen der krankhaften Geschwuelste. Berlin, Germany: Reimer; 1838:56.

(5.) Fiks A. Cystosarcoma phyllodes of the mammary gland-Muller's tumor. Vir chows Arch (Pathol Anat). 1981;392:1-9.

(6.) Lemonaco F. Il tumore filloide della mammella (cistosarcoma filloide di J Muller). Tumori. 1960;46:156-184.

(7.) Lee B, Pack CT. Giant intracanalicular fibromyxoma of the breast. Cancer Res. 1931;15:2583-2595.

(8.) Foote FW, Stewart FW. A histologic classification of carcinoma of the breast. Surgery. 1946;19:74-99.

(9.) Treves N, Sunderland DA. Cystosarcoma phyllodes of the breast: a malignant and a benign tumor. Cancer. 1951;4:1286-1332.

(10.) Jacklin RK, Ridgway PF, Ziprin P, Healy V, Hadjiminas D, Darzi A. Optimising preoperative diagnosis in phyllodes tumor of the breast. J Clin Pathol. 2006;59:454-459.

(11.) Shabalova IP, Chemeris GJ, Ermilova VD, et al. Phyllodes tumor: cytologic and histologic presentation of 22 cases and immunohistochemical demonstration of p53. Cytopathology. 1997;8:177-187.

(12.) Shabb NS. Phyllodes tumor: fine needle aspiration cytology of eight cases. Acta Cytol. 1997;41:321-326.

(13.) Deen SA, McKee GT, Kissin MW. Differential cytologic features of fibroepithelial lesions of the breast. Diagn Cytopathol. 1999;20:53-56.

(14.) Bhattarai S, Kapila K, Verma K. Phyllodes tumor of the breast: a cytohis tologic study of 80 cases. Acta Cytol. 2000;44:790-796.

(15.) Krishnamurthy S, Ashfaq R, Shin HJ, et al. Distinction of phyllode tumor from fibroadenoma: a reappraisal of an old problem. Cancer. 2000;90:342-349.

(16.) Scolyer RA, Mckenzie PR, Achmed D, et al. Can phyllodes tumors of the breast be distinguished from a fibroadenomausing fine needle aspiration cytology? Pathology. 2001;33:437-443.

(17.) Badhe BA, Iyengar KR, Alva N. A study of fibroepithelial tumors of the breast. Indian J Cancer. 2002;39:91-96.

(18.) Jayaram G, Sthaneshwar P. Fine needle aspiration cytology of phyllodes tumors. Diagn Cytopathol. 2002;26:222-227.

(19.) Shimizu K, Korematsu M. Phyllodes tumor of the breast: a cytomorphologic approach based on evaluation of epithelial cluster architecture. Acta Cytol. 2002;46:332-336.

(20.) Dillon MF, Quinn CM, McDermott EW, O'Doherty A, O'Higgins N, Hill ADK. Needle core biopsy in the diagnosis ofphyllodes neoplasm. Surgery.2006; 140:779-784.

(21.) Komenaka IK, El-Tamer M, Pile-Spellman E, Hibshoosh H. Core needle biopsy as a diagnostic tool to differentiate phyllodes tumor from fibroadenoma. Arch Surg. 2003;138:987-990.

(22.) Jacobs TW, Chen YY, Guinee DG, et al. Fibroepithelial lesions with cellular stroma on breast core needle biopsy: are there predictors ofoutcome on surgical excision? Am J Clin Pathol. 2005;124:342-354.

(23.) Lee AHS, Hodi Z, Ellis IO, Elston CW. Histological features useful in the distinction of phyllodes tumor and fibroadenoma on needle core biopsy of the breast. Histopathology. 2007;51:336-344.

(24.) Yohe YS, Yeh IT. "Missed" diagnoses of phyllodes tumor on breast biopsy: pathologic clues to its recognition. Int J Surg Pathol. 2008;16:137-142.

(25.) Foxcroft LM, Evans EB, Porter AJ. Difficulties in the pre-operative diagnosis of phyllodes tumors of the breast: a study of 84 cases. Breast. 2007;16:27-37.

(26.) McGowan TS, Cummings BJ, O'Sullivan B, Catton CN, Miller N, Panzarella T. An analysis of 78 breast sarcoma patients without distant metastases at presentation. Int J Radial Oncol Biol Phys. 2000;46:383-390.

(27.) Lester J, Stout AP. Cystosarcoma phyllodes. Cancer. 1954;7:335-353.

(28.) Oberman HA. Cystosarcoma phyllodes. Cancer. 1965;18:697-710.

(29.) Norris HJ, Taylor HB. Relationship of histologic features to behavior of cystosarcoma phyllodes: an analysis of 94 cases. Cancer. 1967;20:2090-2099.

(30.) Moffat CJ, Pinder SF, Dixon AR, Elston CW, Blamey RW, Ellis IO. Phyllodes tumours of the breast: a clinicopathological review of thirty-two cases. Histopa thology. 1995;27:205-218.

(31.) de Roos WK, Kaye P, Dent DM. Factors leading to local recurrence or death after surgical resection of phyllodes tumor of the breast. Br J Surg. 1999; 86:396-399.

(32.) Kapiris I, Nasiri N, A'Hern R, Healy V, Gui GPH. Outcome and predictive factors of local recurrence and distant metastases following primary surgical treatment of high-grade malignant phyllodes tumors of the breast. Eur J Surg Oncol. 2001;27:723-730.

(33.) Hawkins RE, Schofield JB, Fisher C, Wiltshaw E, McKinna JA. The clinical and histologic criteria that predictmetastases from cystosarcoma phyllodes. Cancer. 1992;69:141-147.

(34.) Barrio AV, Clark BD, Goldberg J, etal. Clinico pathologic features and longterm outcomes of 293 phyllodes tumors of the breast. Ann Surg Oncol. 2007;14: 2961-2970.

(35.) Grabowski J, Salzstein SL, Sadler GR, Blair SL. Malignant phyllodes tumors: a review of 752 cases. Am Surg. 2007;73:967-969.

(36.) Kleer CG, Giordano TJ, Braun T, Oberman HA. Pathologic, immunohistochemical and molecular features of benign and malignant phyllodes tumors of the breast. Mod Pathol. 2001;14:185-190.

(37.) Feakins RM, Mulcahy HE, Nickols CD, Wells CA. p53 expression in phyllodes tumors is associated with histological features of malignancy but does not predict outcome. Histopathology. 1999;35:162-169.

(38.) Yonemori K, Hasegawa T, Shimizu C, et al. Correlation of p53 and MIB-1 expression with both the systemic recurrence and survival in cases of phyllodes tumors of the breast. Pathol Res Pract. 2006;202:705-712.

(39.) Sawyer EJ, Poulsom R, Hunt FT, et al. Malignant phyllodes tumours show stromal overexpression of c-myc and c-kit. J Pathol. 2003;200:59-64.

(40.) Tan PH, Jayabaskar T, Yip G, et al. p53 and c-kit (CD117) protein expression as prognostic indicators in breastphyllodes tumors: a tissue microarray study. Mod Pathol. 2005;18:1527-1534.

(41.) Tse GMK, Tsang AKH, Putti TC, et al. Stromal CD10 expression in mammary fibroadenomas and phyllodes tumors. J Clin Pathol. 2005;58:185-189.

(42.) Lee AHS. Recentdevelopments in the histological diagnosis of spindle-cell carcinoma, fibromatoses and phyllodes tumor of the breast. Histopathology. 2008;52:45-57.

(43.) Lerwill MF. Biphasic lesions of the breast. Sem Diagn Pathol.2004;21:4856.

(44.) Layfield LJ, Hart J, Neuwrith H, Bohman R, Trumbull WE, Giulionao AE. Relation between DNA ploidy and the clinical behavior ofphyllodes tumos. Cancer. 1989;64:1486-1489.

(45.) Lae M, Vincent-Soloman A, Savignoni A, et al. Phyllodes tumors of the breast segregate in two groups according to genetic criteria. Mod Pathol. 2007; 20:435-444.

(46.) Jones AM, Mitter R, Springall R, et al. A comprehensive genetic profile of phyllodes tumours ofthebreastdetects importantmutations, intra-tumoral genetic heterogeneity and new genetic changes on recurrence. JPathol. 2008;214:533 544.

Dilip Giri, MD

From the Department of Pathology, Memorial Sloan-Kettering Cancer Center, New York, NY.

The author has no relevant financial interest in the products or companies described in this article.

Presented in part at The Surgical Pathology of Neoplastic Diseases course, Memorial Sloan-Kettering Cancer Center, New York, New York, May 12-16, 2008.

Reprints: Dilip Giri, MD, Department of Pathology, Memorial SloanKettering Cancer Center, 1275 York Avenue, New York, NY 10065 (e-mail: girid@mskcc.org).
Morphologic Features of Phyllodes Tumor (PT), Per World Health
Organization Classification (3)

                           Benign PT
Stromal hypercellularity   Modest
Cellular pleomorphism      Little
Mitosis                    Few, if any

Margins                    Well circumscribed (pushing)
Stromal pattern            Uniform stromal distribution

Heterologous stromal       Rare
differentiation
Overall average            60
distribution

                           Borderline PT

Stromal hypercellularity   Modest
Cellular pleomorphism      Moderate
Mitosis                    Intermediate

Margins                    Intermediate
Stromal pattern            Heterogeneous stromal
                            distribution
Heterologous stromal       Rare
differentiation
Overall average            20
distribution
                           Malignant PT

Stromal hypercellularity   Marked
Cellular pleomorphism      Marked
Mitosis                    Numerous (>10 per 10 high-
                           power fields)
Margins                    Invasive
Stromal pattern            Marked stromal overgrowth

Heterologous stromal       Not uncommon
differentiation
Overall average            20
distribution
Gale Copyright: Copyright 2009 Gale, Cengage Learning. All rights reserved.