Fibroepithelial Lesions in the Breast of Adolescent Females: A Clinicopathological Study of 54 Cases

Abstract

Background

Fibroepithelial lesions (FELs) are the most frequent breast tumors in adolescent females. The pubertal hormonal surge could impact the growth and microscopic appearance of FELs in this age group. In this study, we evaluate the morphology and clinical behavior of FELs in adolescents.

Design

We searched the 1992–2012 pathology database for FELs in females 18 years old or younger (F≤18y). Seven FELs from 1975–83 were also included. Three pathologists reviewed all available material. Patient (pt) characteristics and follow-up information were obtained from electronic medical records.

Results

Forty-eight F≤18y had 54 FELs with available slides. Thirty (67%) pts were Caucasian, 12 (27%) African-American, two (4%) Hispanic, one (2%) Asian; three were of unknown race/ethnicity. Median age at diagnosis was 16 y. Median age at menarche was 12 y; most (96%) FELs occurred after menarche (median interval 48 months). All pts underwent lumpectomy; one required subsequent mastectomy. The FELs were 34 fibroadenomas (FAs) (11 usual, 23 juvenile), and 20 phyllodes tumors (PTs) (16 benign, one borderline and three malignant). Eight (35%) juvenile FAs showed slight intratumoral heterogeneity. The mean mitotic rate was 1.3 mitoses/10 high-power fields (HPFs) (range, 0–6) in usual FAs, 1.8/10 HPFs in juvenile FAs, 3.1/10 HPFs in benign PTs, 10/10 HPFs in the borderline PT and 17/10 HPFs in malignant PTs. The mean follow-up for 29 pts with 33 FELs was 44 months. Two (10%) PTs recurred locally (a benign PT at 18 months, and a borderline PT at 11 months). Both recurrent PTs had microscopic margins <1 mm.

Conclusion

Mitotic activity in FAs from adolescents can be substantial and this finding should be interpreted cautiously. Awareness of the morphologic features of FELs in adolescents is important to avoid overdiagnosis of PTs, which can lead to additional unnecessary and potentially disfiguring surgery.

Introduction

Breast lesions in children and adolescents are rare, and include benign and malignant tumors, ranging from fibroadenoma (FA) to phyllodes tumor (PT) and secretory carcinoma. Systemic diseases such as leukemia and lymphoma, and metastases from an extra-mammary malignancy also occasionally involve the juvenile breast. Of all these tumors, fibroepithelial lesions (FELs) are by far the most common in females 18 years old and younger (F<18y) (1, 2). The varied morphologic spectrum of FELs is reflected by the various descriptive terms such as fetal, cellular, juvenile, giant and variant FA, tubular adenoma, hamartoma, cystosarcoma phyllodes and PT that have been used to designate these lesions in the past (3–7). Very few series of FELs in young women have been reported in the literature, and our knowledge of the morphologic spectrum of FELs in the adolescent breast remains limited. In particular, it is unclear if the criteria used for the diagnosis of FELs in the adult female breast (including tumor size, stromal cellularity and mitotic activity) carry the same weight in FELs from adolescent girls, as in adolescence the breast undergoes rapid and substantial alterations during the course of puberty. Overdiagnosis of FELs in adolescents could lead to unnecessary and potentially disfiguring surgery due to disruption of the developing breast bud.

To address these questions, we investigated the morphology and clinical behavior of FELs in adolescent females.

Materials and Methods

Case Selection

After obtaining IRB approval, we searched the digital records of our surgical pathology service for breast specimens obtained from patients 18 years old or younger treated at our institution between 1992 to 2012. We also included in our series seven additional patients in the same age group who were treated at our center for a FEL between 1975 and 1983, part of a previously published series (8). Our study population consists of a total of 48 patients with an index diagnosis of a FEL and slides available for review. A total of 54 FELs were examined, including 50 FELs surgically excised at our center and four FELs excised at other hospitals but with slides available for review in our consultation files and available clinical follow-up. One of the four FELs originally excised at an outside institution was submitted due to a diagnostic question, while the other three were reviewed because the patients transferred their care to our institution. Information on the gross size of the FEL was extracted from the original pathology report. All pertinent and available slides, including slides of ipsilateral breast needle core biopsies, excisional biopsies and/or mastectomy specimens were retrieved and reviewed by three breast pathologists, two of whom (EB and DDG) were unaware of the FEL classification in the original pathologic diagnosis. We recorded the microscopic size of the FEL, its borders and growth pattern, presence of stromal overgrowth (defined as the absence of any epithelial component at 40× final magnification [10× ocular piece and 4× objective]), stromal cellularity, nuclear atypia, final margin status and epithelial hyperplasia. Mitotic activity of the stromal component was assessed by counting mitoses per 10 high-power fields (HPFs) at 400× final magnification (10× ocular piece and 40× objective) in the most cellular and mitotically active areas of the lesion. FELs were classified into different categories based upon a combination of morphologic criteria, including those criteria specified in the WHO 2012 classification (9). The categories of classification in this study include: usual/adult FA, juvenile FA (including variant juvenile FA), benign PT, borderline PT and malignant PT.

Information regarding clinical presentation, radiologic findings of the lesions and patient follow-up was extracted from the electronic medical records. Whenever available, we also collected information on the patient’s age at diagnosis of the index FEL, age at menarche, time from menarche to diagnosis of FEL, race or ethnicity, and personal/family medical history.

Results

Clinical Characteristics

Our study cohort consists of 48 F≤18 y who developed 54 FELs. The clinical characteristics of the 48 patients are summarized in Table 1. Forty-three patients (43/48; 90%) had a solitary FEL; five patients had multiple FELs, including one patient with three ipsilateral FELs, two patients with two ipsilateral FELs and two other patients with bilateral FELs (Table 2). Thirty-three FELs occurred in the right breast and 21 in the left breast. Information on race or ethnicity was available for 45 patients: 30/45 (67%) were Caucasian, 12/45 (27%) African-American, 2/45 (4%) Hispanic and 1/45 (2%) Asian. Age at presentation ranged from 10 to 18 y (mean 15.4 y, median 16 y). Mean and median age at menarche was 12 y (range, 10–14) for 27 pts with available information. In most patients (26/27; 96%) the index FEL was diagnosed at a median of 48 months (range, 0–72) after menarche (mean 45 months). Only one patient developed a juvenile FA 12 months prior to menarche. No information on body mass index or exogenous hormonal treatment was available. The phase of the menstrual cycle at the time of surgery was also unknown.

Table 1.

Clinical Characteristics of 48 Adolescent Females with 54 Fibroepithelial Lesions

Age (y)
Mean	15.4
Median	16
Range	10–18
Breast Laterality (N)
Right	33
Left	21
Race/Ethnicity (known for 45 pts)
Caucasian	30
African-American	12
Hispanic	2
Asian	1
Unknown	3
Median Age at Menarche (y) (known for 27 pts)	12 (10–14)
Median Time from Menarche to Diagnosis (mo)*	48
Presentation (N) (known for 42 FELs)
Palpable Mass	41
Ultrasound†	1
Number of lesions
Solitary lesion (N)	43
Multiple lesions (N)
3 Ipsilateral	1
2 Ipsilateral	2
Bilateral	2

^*1 pt presented 12 months prior to menarche

^†Pt undergoing ultrasound for ipsilateral FEL

y = years; N = number; mo= months

Table 2.

Patients with Multiple Fibroepithelial Lesions

Patient	Laterality	Fibroepithelial Lesions
1	Ipsilateral	3 juvenile fibroadenomas
2	Ipsilateral	1 fibroadenoma, 1 juvenile fibroadenoma
3	Ipsilateral	2 juvenile fibroadenomas
4	Bilateral	1 malignant phyllodes tumor, 1 juvenile fibroadenoma
5	Bilateral	2 juvenile fibroadenomas

Information on clinical presentation was available for 43 FELs. Most (42/43; 98%) presented as a palpable mass. The non-palpable FEL was noted during sonographic evaluation of a palpable ipsilateral FEL. Two patients reported rapid growth of the breast mass; the corresponding FELs were a malignant PT spanning 25 cm in a 15-year-old patient, and a juvenile FA measuring 3 cm in a 14-year-old patient. In three patients the FEL was described by the surgeon as filling almost the entire breast (75% to 100%); the corresponding FELs were a benign PT spanning 13 cm (11-year-old patient), a juvenile FA variant measuring 7 cm (13-year-old patient), and a juvenile FA spanning 5 cm (11-year-old patient). An 18 y patient delivered a child 11 months prior to diagnosis of the index FEL. She had reported that the lesion increased in size during pregnancy. At surgical excision her FEL was a 2.5 cm usual FA. None of the F≤18y had a personal history of malignancy. Thirteen patients had a family history of breast carcinoma in first or second-degree relatives.

All patients underwent surgical excision of the FEL; one patient with a malignant PT required subsequent mastectomy because the lesion involved the surgical margins. Eight patients underwent initial diagnostic sampling of the FEL by needle core biopsy and one by fine needle aspiration.

Pathologic Characteristics

Of the 54 FELs, 34 were classified on re-review as FAs (11 usual/adult type FAs, 23 juvenile FAs), and 20 as PTs (16 benign, one borderline and three malignant). The clinical features associated with the 54 FELs are outlined in Table 3. Table 4 summarizes the classification and pathologic characteristics of the 54 FELs in the study. The mean size of all FELs was 3.4 cm (range, 0.5–25): the mean size of FAs was 2.9 cm (range, 0.5–7) and that of PTs was 6.3 cm (range, 1–25). The tumors were excised with minimal surrounding breast tissue, as demonstrated in Figure 1. Figure 2 shows the frequency of each FEL subtype in the group of 47 lesions diagnosed between 1992 and 2012.

Table 3.

Clinical Characteristics of 54 Fibroepithelial Lesions in Females 18 Years of Age and Younger

	Fibroadenoma		Phyllodes Tumor
	Usual/Adult	Juvenile	Benign	Borderline	Malignant
N	11	23	16	1	3
Age (y)
Mean	17.1	15	15	15	15
Median	17	16	15	15	15
Range	15–18	11–18	10–18	15	13–16
Laterality (N)
Right	6	15	9	0	3
Left	5	8	7	1	0
Race/Ethnicity *
Caucasian	9	12	8	1	2
African-   American	2	7	5	0	1
Hispanic	0	1	1	0	0
Asian	0	0	1	0	0
Mean Age at Menarche (y) †	12 (6 pts)	12 (14 pts)	12 (7 pts)	N/A	N/A
Median Time from Menarche to Diagnosis (mo)£	72 (6 pts)	36 (14 pts)	42 (7 pts)	N/A	N/A
Presentation ₳ (N)
Palpable   Mass	10	19	11	N/A	2
US¥	1	0	0	N/A	0

^*Information available for 45 patients (50 FELs)

^†Information available for 27 patients

^£1 pt with a Juvenile FA variant presented 12 mo prior to menarche

^₳Information available for 43 FELs

^¥Pt undergoing ultrasound evaluation for ipsilateral FEL

N = number; y = years; mo = months; N/A = not available

Table 4.

Pathologic Characteristics of 54 Fibroepithelial Lesions in Females 18 Years of Age and Younger

	N	Mean Size (cm)	Mean Mitotic Count / 10 HPF	Borders		Growth Pattern		Epithelial Hyperplasia
				Circumscribed	Infiltrative	Intracanalicular	Pericanalicular
Fibroadenoma	34	2.9	1.6	34	0	10	24	9
Usual	11	2.6 (0.7– 4.5)	1.3 (0–6)	11	0	10	1	2 (18%)
Juvenile	23	3.1 (0.5– 7)	1.8 (0–7)	23	0	0	23	7 (30%)
Phyllodes Tumor	20	6.3	5.6	12*	6*	14	6	8
Benign	16	4.9 (1– 13)	3.1 (1–7)	12	3	11	5	8 (50%)
Borderline	1	N/A	10	0	1	0	1	0
Malignant	3	14.5 † (4, 25)	17 (12–20)	0	2	3	0	0

^*Borders not assessable in 1 benign PT and 1 malignant PT

^†Gross size available for 2 lesions, which measured 4 cm and 25 cm

N/A = Size not available for the borderline PT

Figure 1.

Gross Picture of Benign Phyllodes Tumor. Cut section shows bulging tissue with frond-like excrescences.

Figure 2.

Classification of 47 fibroepithelial lesions in our Department of Pathology from 1992 to 2012 in females 18 years of age and younger.

Eight FELs (six BPTs and two juvenile FAs on excision) underwent an initial needle core biopsy. For the six BPTs, in five cases the histologic findings on biopsy raised the possibility of a PT (i.e. fibroepithelial lesion with increased stromal cellularity); the remaining case was diagnosed as a FA histologically on biopsy, but the clinical/radiological findings raised the possibility of a PT. For the two juvenile FAs, the initial core biopsies favored a diagnosis of juvenile FA.

Fibroadenomas

Usual/Adult Fibroadenomas

There were 11 usual FAs with morphology indistinguishable from that of usual FAs commonly seen in adult women. The usual FAs amounted to 20% of all our cases and constituted 23% of the 47 available FELs diagnosed between 1992 and 2012. The lesions had circumscribed borders, and showed even distribution of glands and stroma. All but one usual-type FA displayed an intracanalicular growth pattern (Figure 3 A–C), and none had stromal overgrowth. Slightly increased stromal cellularity was noted in 27% of cases, but no nuclear atypia was identified. The most mitotically active FA had 6 mitoses/10HPFs and was from the breast of the 18- year-old girl who had given birth 11 months prior to excision of the index lesion. Her FA also showed very focal lactational changes. The stromal component including this case with lactational change had a mean mitotic count of 1.3/10 HPFs (range, 0–6). If this case is excluded, as the recent pregnancy and possible lactation could have affected some of the tumor characteristics, the mean mitotic count of usual FAs was 0.8 mitoses/10 HPFs (range, 0–3). Epithelial hyperplasia was present in two cases.

Figure 3.

A–C) Usual fibroadenoma with an intracanalicular growth pattern (10×, 100×, 400×). The arrow points to a mitotic figure. D–F) Juvenile fibroadenoma with a pericanalicular growth pattern (10× 100×, 400×). G–I) Benign phyllodes tumor (10×, 100×, 400×). J–L) Malignant phyllodes tumor with liposarcomatous areas (10×, 100×, 400×).

Juvenile Fibroadenomas

Twenty-three FELs were so called “juvenile” FAs. They represented 43% of all cases and 45% of the 47 FELs diagnosed between 1992 and 2012. All juvenile FAs showed circumscribed borders and had no stromal overgrowth. These tumors showed a pericanalicular growth pattern, and the glandular component was well-developed and retained a lobular architecture (Figure 3 D–F). Focally, increased stromal cellularity was seen in 61% of cases but none of the cases showed nuclear atypia. The mean mitotic count was 1.8 mitoses/10 HPFs (range, 0–7).

Juvenile FAs were characterized by a somewhat uniform quality of the stroma, which showed no overt distinction between intra- and extra-lobular stroma, and fairly even cellularity throughout the lesion. Short collagen bundles regularly admixed with the stromal cells were also a common finding (Figure 3E). Micropapillary epithelial hyperplasia was present in 30% of juvenile FAs. The combination of these morphologic features allowed reproducible diagnosis of juvenile FA in this series.

Eight juvenile FAs showed slight morphologic variation. Just as juvenile FAs, these tumors had well-circumscribed borders, pericanalicular growth pattern and uniformly cellular and collagenous stroma. However, they also showed focal areas of stromal expansion throughout the lesion (Figure 4). We regarded these tumors as variant forms of juvenile FA, as the morphology closely resembled that illustrated by Azzopardi et al. (5). Four of the eight cases with this variant pattern also showed close juxtaposition to gland-rich areas with tubular arrangement, similar to those seen in juvenile FAs, and areas with more cellular and collagenous stromal component, characteristic of the juvenile FA variant pattern. As a result, these variants of juvenile FA showed some degree of intratumoral heterogeneity (Figure 5) that could raise the differential diagnosis of PT; however, no nuclear atypia was identified in any of these cases. One patient with a variant juvenile FA presented 12 months prior to menarche. The age of menarche was available for a total of 14 patients with a juvenile FA and the interval between menarche and the development of the lesion was the same year to 60 months.

Figure 4.

Juvenile fibroadenoma variant. A) Well-circumscribed borders at low magnification. B) Pericanalicular growth pattern with collagenized and uniformly cellular stroma.

Figure 5.

Juvenile fibroadenoma variant. A) Intratumoral heterogeneity with variable epithelial to stromal ratio (10×). B) Border between compact glands, similar to juvenile fibroadenoma, adjacent to an area with a cellular and collagenous stromal component, similar to the variant pattern (100×). C) No nuclear atypia was identified in these cases (200×).

Phyllodes Tumors

Benign Phyllodes Tumors

Sixteen FELs were benign PTs. They represented 30% of all cases and 28% of the 47 available FELs diagnosed between 1992 and 2012. Increased stromal cellularity and mild to moderate nuclear atypia were identified in all 16 cases (Figure 3 G–I). No stromal overgrowth was present. The mean mitotic count was 3.1 (range, 1–7). Half of the cases showed epithelial hyperplasia.

Borderline and Malignant Phyllodes Tumors

There were one borderline and three malignant PTs, representing 7% of all cases and 4% of the 47 available FELs diagnosed between 1992 and 2012. Information on the tumor gross size was available only for two malignant PTs that measured 4 cm and 25 cm, respectively. Tumor borders were infiltrative in two cases and could not be assessed in the other two. Increased stromal cellularity was present in all four cases and the malignant PTs also showed stromal overgrowth. Nuclear atypia was moderate in the borderline PT and marked in the malignant PTs. One of the three malignant PTs had liposarcomatous areas (Figure 3 J–L). The mitotic count for the borderline PT was 10/10 HPFs. The mean mitotic count for the malignant PTs was 17/10 HPFs (range, 12–20).

Surgical Management and Margin Assessment

All FELs underwent surgical excision. Surgical margins could be evaluated in 40/54 (74%) cases. The final inked surgical margin was positive in 24 cases: 18 FAs (five usual FAs and 13 juvenile FAs, five of which were variant juvenile FAs), and six benign PTs. The FEL was close (≤ 1 mm) to the final inked margin in 12 cases: eight FAs (four usual FAs and four juvenile FAs, two of which were variant juvenile FAs), three benign PTs and one borderline PT. One patient with a malignant PT underwent mastectomy after two consecutive excisions yielded positive margins. Review of the surgical operative notes from the patients in this study indicates that surgeons made conscious efforts in preserving uninvolved breast tissue by separating the mass by sharp and blunt dissection.

Clinical Follow-Up

Follow-up information was available for 29 patients with 33 FELs (seven FA, 17 juvenile FA, six benign PTs, one borderline PT, two malignant PTs). The mean overall follow-up time was 44 months (range, 1–278). Two patients, one with a benign PT and one with a borderline PT developed local recurrence with morphology similar to that of the index lesion (Table 5). In both cases the original tumor had been excised with a less than 1 mm margin; the index BPT extended to ink and the index borderline PT was less than 1 mm to the ink. None of the patients with available follow-up developed metastatic disease.

Table 5.

Information for 2 Recurrent Cases

	Case 1	Case 2
Diagnosis	Benign Phyllodes Tumor	Borderline Phyllodes Tumor
Age at Diagnosis (y)	10	15
Race/Ethnicity	Hispanic	Caucasian
Months to Recurrence	18	11
Gross Size (cm)
Index Tumor	4.5	N/A
Recurrent Tumor	8.5	N/A
Borders
Index Tumor	Infiltrative	Infiltrative
Recurrent Tumor	Infiltrative	Infiltrative
Increased Stromal Cellularity	Yes	Yes
Index Tumor	Yes	Yes
Recurrent Tumor
Mitotic Count per 10 HPF
Index Tumor	5	10
Recurrent Tumor	4	10
Stromal Nuclear Atypia
Index Tumor	Mild	Moderate
Recurrent Tumor	Mild	Moderate
Margin Status
Index Tumor	Tumor at ink	< 1 mm to ink
Recurrent Tumor	<1 mm to ink	Negative (clearance > 1 mm to tissue edge)

N/A = Size not available for the borderline PT

y = years; HPF = high-power fields

Discussion

Fibroepithelial lesions are the most frequent tumors in the adolescent breast (2, 10). FELs are reportedly more frequent in African-American girls (11) but geographic factors may also be important. In our series, nearly 70% of our patients were Caucasian, and African-American girls were the second most common ethnic group.

Most reports of FELs in adolescents consist of sporadic cases, and only very few series have been published (6),(12),(13),(14),(15),(16),(4),(17). The majority of FELs in this age group are benign, and consist of FAs with usual/adult morphology or of the so called “juvenile” type. In usual/adult type FAs, stroma and epithelium are evenly distributed throughout the tumor. Stromal cellularity is low and hyalinization uncommon.

FAs characterized by more cellular stroma and increased number of glands have been referred to as “juvenile” but they are not exclusive of adolescents. In 1971 Ashikari et al. (6) studied 181 FAs in adolescent females ranging in age from 10 to 20 y. Twelve of these tumors were classified as juvenile FAs. The authors described juvenile FAs as floridly glandular, with cellular and dense stroma, and larger in size compared to FAs of adult type (6). In 1974, Nambiar and Kutty (16) reported the findings in 25 adolescent females ranging in age from 11 to 20 y that had giant FA of the breast, with an average size of 12 cm. Microscopically the giant FAs consisted of hyperplastic and cellular stroma separating slit-like canaliculi, clefts and cysts. Focal epithelial hyperplasia with bud-like outgrowths was also noted (16). Some of these cases might have represented examples of florid pseudoangiomatous hyperplasia. In 1985, Pike and Oberman (4) reported a series of 25 patients ranging in age from 11 to 19 y with “juvenile (cellular) adenofibromas.” Fourteen of the 25 tumors were characterized by prominent stromal cellularity with pericanalicular arrangement around the glands. Fourteen tumors also displayed noticeable intercellular stromal collagen that was prominent in five cases (4).

Although we acknowledge that the term “juvenile” has limitations, we used it in this study for easier reference, to emphasize the somewhat peculiar morphology of these lesions rather than introduce a different terminology. Even though juvenile FAs are relatively rare in common practice, they stood out as a distinct group in our systematic review of FELs in adolescents. Juvenile FAs in our series were also characterized by increased stromal cellularity and collagenized stroma. Although the growth pattern of the stroma in juvenile FAs is usually described as pericanalicular, it is more accurate to say that these tumors contain a substantial glandular component, which often displays a somewhat lobular architecture. At low magnification these lesions may resemble an adenosis tumor or tubular adenoma, but on closer examination they show expanded stroma composed of bland myofibroblasts in a fascicular arrangement, admixed with short and inconspicuous collagen fibers.

Eight of the 23 juvenile FAs in our series showed slight variation in morphology, reminiscent of the “fibroadenoma variant” described by Azzopardi et al. as “a fibroepithelial lesion composed of collagenous and cellular stroma that is related to a fibroadenoma in structure and composition but shows sufficient morphologic variation to raise problems in precise classification” (5). Some variant FAs in our series showed slight intratumoral heterogeneity, raising the differential diagnosis of a benign PT. However, no nuclear atypia was identified in any of these lesions and the stroma, albeit cellular, maintained uniform and collagenized fibrous quality throughout the tumor, including the areas with a relatively sparser glandular component. In our study, pericanalicular growth pattern, uniform quality of the stroma with no identifiable separation between intra-lobular/periglandular and extra-lobular stromal areas, fascicular growth of the myofibroblasts composing the lesion, presence of short collagen fibers, and lack of nuclear atypia throughout the lesion consistently characterized all variant juvenile FAs. Close attention to these morphologic features will help pathologists to recognize these unusual tumors and avoid an overdiagnosis of benign PT. Just as for other FELs in this age group, emphasis on mitotic activity alone could be misleading in the evaluation of these tumors, as juvenile FA, including variant juvenile FAs, can display considerable mitotic activity in this young age group.

Hamartomas have been reported in adolescents. They are characterized by a disproportionate distribution of normal mammary tissue components, including adipocytes. The morphological disarray characteristic of hamartomas contrasts with the somewhat “organized” architecture of the “juvenile” FAs in our series, and we did not consider the diagnosis of hamartoma for any of our cases, although some tumors in reported series of hamartomas (18–20) show some similarities with the current lesions. Although we do not favor this hypothesis, the possibility that a “juvenile” FA may be a variant of breast hamartoma cannot entirely be ruled out.

The term cystosarcoma phyllodes was introduced in 1838 by Johannes Muller to designate large nodular stromal tumors consisting of leaf-like fronds projecting into cystic spaces. Two large series of (cystosarcoma) phyllodes tumors were published in the 1950s by Treves and Sunderland (21) and Lester and Stout (22). For the reader’s ease, in this discussion we use the term PT also for lesions that had been referred to as cystosarcoma phyllodes in the original publication. Initially pathologists grouped all large fibroepithelial tumors with hypercellular stroma into this malignant category until it was recognized that “size is a secondary consideration, represents the end result of prolonged or rapid growth, and is not an essential consideration in making the diagnosis of cystosarcoma phyllodes”(21). In 1970, Amerson (12) reported PTs in seven patients between the ages of 10–17 y. Diagnostic criteria included presence of epithelial hyperplasia and a greater degree of stromal hyperplasia compared to that usually seen in FAs. Giant FAs were excluded. Cellular atypia, mitotic activity and tumor margin were used to try and separate lesions as benign versus malignant PTs. In 1998 Rajan et al. (13) examined PTs in 45 patients ranging in age from 10–24 y. PTs were categorized based on stromal overgrowth, invasive tumor borders, cytologic atypia, stromal mitoses, and necrosis. Thirty-four cases were classified as benign PTs, eight as low-grade malignant, and three as high-grade malignant. Follow-up information was available for 36 patients for a mean of 58.4 months. Six patients developed local recurrence, including four with benign PTs and two with malignant PTs. At the last follow-up, 34 patients were alive with no evidence of disease, one patient was alive with a pulmonary metastases, and one patient died of unrelated causes (13). Nowadays, the preferred term for these lesions is phyllodes tumors (PTs), as cystic growth is not a characteristic feature, and the percentage of cases with malignant clinical behavior is relatively low overall. The WHO 2012 Classification of Tumours of the Breast classifies PTs as benign, borderline, and malignant based on the assessment of six morphologic features (tumor border, stromal cellularity, stromal atypia, mitotic activity, stromal overgrowth, and malignant heterologous elements) (9).

PTs often grow rapidly to a large size in a short time. Mitotic activity is one of the morphologic features used in the classification of PT. It is an expression of the biology of the tumor, but in mammary FELs arising in adolescent females it may be also affected by hormonal stimulation. In particular, the stromal cells of FELs express the progesterone receptor. There are higher levels of progesterone receptor A and B in FAs compared to normal breast tissue (23) and higher levels of progesterone have been detected in benign PTs compared to malignant PTs (24). The rapid growth of some FELs in adolescents may be in part related to the hormonal surge characteristic of puberty. A notable finding in our series is a relatively high mitotic rate in all FELs in adolescent females, including FAs, that could raise concern and lead to misclassification of the tumor as a PT. As demonstrated by overall benign clinical follow-up in our series, FAs with substantial mitoses did not correlate with worse behavior.

Margin clearance for all histologic subtypes of PTs has been examined in the literature, however the number of adolescent patients in published series is generally underrepresented. Most authors recommend a margin of at least 1 cm for PTs in the general population based on data suggesting that margin status is a key variable that correlates with recurrence (25),(26),(27),(28),(29),(30),(31). Teo et al. (28) examined the local recurrence rates specifically in young Asian patients with PTs. The study involved a retrospective review of 44 patients with 45 tumors (42 benign, 2 borderline, 1 malignant) aged 15–25 y at the time of diagnosis with a mean and median age of 20.7 y. The margin status was clear for 11 cases, involved in 15 cases and unknown in 19 cases. All 15 patients with involved margins had benign PTs. No recurrences were reported with a total follow-up period of 2093 months. It was concluded that in this patient population, simple excision, followed by close follow-up is sufficient, regardless of surgical margin status (28).

The recurrence rate of PT correlates with tumor grade and is lowest for benign PTs, ranging around 11% in a large series (32). A nomogram has been developed by Singapore General Hospital to predict the recurrence-free survival of patients with PTs. The nomogram was developed based on data from patients with PTs ranging in age from 15–79 y (mean age 42 y). 605 PTs were included in the study (72.7% benign, 18.4% borderline and 8.9% malignant). Stromal atypia, overgrowth and surgical margins were found to be predictive factors for PT recurrence, with mitoses achieving near significance and these are the four parameters used in the nomogram (32). Although the nomogram constitutes an excellent tool to estimate recurrence free survival, it likely included only a limited number of patients aged 18 years old or younger, and cannot be specifically applied to predict the behavior of FELs arising in pubertal patients.

In summary, in our series, fibroepithelial lesions in adolescents were benign in two-thirds (34/54, 63%) of cases and consisted of juvenile fibroadenomas in nearly half of the cases and usual fibroadenoma in nearly one-fourth. Of note, the seven FELs from patients that were part of a previously published study (8) were classified on re-review as two juvenile FAs, three BPTs, one borderline PT and one malignant PT in this study. We found that mitotic activity in the stroma of FELs from females 18 years old and younger is common, and can even be substantial. It is important to be aware of this finding because increased mitotic activity is one of the six morphologic parameters used in the classification of PTs in adults. Except for mitotic activity, usual FAs and benign and malignant PTs are morphologically indistinguishable from their adult counterparts. Despite the cellular nature of FELs occurring in adolescent females, our results show benign prognosis for these tumors consistent with prior studies, and do not reveal a predisposition to the subsequent development of a PT or carcinoma in either breast (10),(4),(6),(12),(33). “Juvenile” FA, the most common FEL lesion in our series, is characterized by a floridly glandular lobular architecture, collagenous stroma and pericanalicular growth pattern. Juvenile FA with variant morphology, as first described by Azzopardi et al. (5) may show slight stromal expansion and intratumoral heterogeneity but the stromal cells lack nuclear atypia. Awareness of these morphologic features is important to avoid overdiagnosis of PTs in adolescents, which can lead to additional surgery and carries the risk of disrupting the developing breast bud.

These results provide a useful reference to pathologists and clinicians treating adolescents with FELs.