Article Text

Download PDFPDF
Benign spindle cell lesions of the breast: a diagnostic approach to solitary fibrous tumour, nodular pseudoangiomatous stromal hyperplasia and nodular fasciitis
  1. Julia R Naso1,
  2. Connie G Chiu2,
  3. Michelle E Goecke2,
  4. Debra Chang3,
  5. Carolyn J Shiau1,4
  1. 1 Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, BC, Canada
  2. 2 Department of Surgery, Royal Columbian Hospital, New Westminster, BC, Canada
  3. 3 Department of Radiology, Royal Columbian Hospital, New Westminster, BC, Canada
  4. 4 Department of Pathology, Royal Columbian Hospital, New Westminster, BC, Canada
  1. Correspondence to Dr Carolyn J Shiau, Department of Pathology, Royal Columbian Hospital, New Westminster, BC, Canada; carolyn.shiau{at}


Benign spindle cell lesions of the breast include neoplastic and reactive entities that are diagnostically challenging given their rarity and similar histomorphology. Accurate diagnosis on percutaneous core biopsy within this category is essential as some lesions require excision and surveillance, whereas others may be observed. We present three cases of rare benign spindle cell lesions of the breast that reflect the diversity of this group: solitary fibrous tumour, nodular pseudoangiomatous stromal hyperplasia and nodular fasciitis. Through these cases, we discuss the associated differential diagnosis and demonstrate how emerging ancillary studies can be integrated into a diagnostic approach. We highlight distinctive clinical and histopathological features and summarise recent updates to the clinical management of these lesions. An organised approach to the broad differential of spindle cell lesions is essential for appropriate diagnosis and treatment.

  • breast
  • soft tissue tumours
  • diagnosis
  • immunohistochemistry

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Benign spindle cell lesions of the breast are a diverse group of entities sharing a similar histopathological appearance. They may appear clinically and radiographically similar to common benign lesions (eg, fibroadenomas) or may appear falsely malignant. Management differs greatly among types of benign spindle cell lesions, ranging from observation to complete excision with long-term follow-up. Accurate diagnosis on percutaneous core biopsy is thus imperative for the optimal management of these lesions.

Benign spindle cell lesions are often challenging to diagnose because of the breadth of the differential diagnosis, the overlap in histomorphology between diagnostic entities and the rarity of these lesions in the breast. The most diagnostically challenging entities in the differential diagnosis for bland-appearing spindle cell lesions are described in table 1. These include both benign and malignant entities, which are critical to distinguish. Less challenging entities include leiomyomas, neurofibromas, schwannomas and myxomas, which rarely occur in the breast but are common in other sites.1 Careful morphological assessment together with an initial immunohistochemistry panel including p63 and pancytokeratin (to exclude spindle cell carcinoma1) are recommended. Inclusion of smooth muscle actin (SMA) and CD34 in the initial panel can stratify the remaining lesions in the differential (table 1).

Table 1

Differential diagnosis for diagnostically challenging bland-appearing spindle cell lesions of the breast.

We present our diagnostic approach to three patients with rare spindle cell lesions of the breast: solitary fibrous tumour (SFT), nodular pseudoangiomatous stromal hyperplasia (PASH) and nodular fasciitis (NF). These cases facilitate discussion of the differential diagnosis and illustrate the clinical diversity of benign spindle cell lesions of the breast.

Case 1

A 52-year-old woman presented with an 11 mm non-tender, mobile, round breast mass identified on routine mammography. Ultrasound identified a well-circumscribed, hypoechoic solid lesion with internal vascularity (figure 1A). The lesion grew 2 mm over 21 months, prompting a core biopsy. The biopsy revealed densely packed haphazardly arranged spindle cells with mild nuclear pleomorphism and background staghorn-type vessels (figure 1B). Neither mitotic activity nor necrosis was identified. Immunohistochemistry showed that the lesion was positive for CD34 (figure 1C) and signal transducer and activator of transcription 6 (STAT6) (figure 1D), and was negative for SMA (figure 1E), p63 and CK5/6. These findings were compatible with SFT and reviewed at a multidisciplinary conference. A chest CT scan was performed to exclude a metastatic focus of SFT. The lesion was completely excised (figure 1F), and final pathology confirmed that the lesion was a SFT with benign histological features. Surveillance with yearly diagnostic mammography was arranged.

Figure 1

Solitary fibrous tumour. (A) The lesion appears as a discrete hypoechoic nodule on ultrasound. (B) On core biopsy, the nodule contains densely packed spindle cells with staghorn-type vessels (H&E). On immunohistochemistry, the spindle cells are positive for (C) CD34 and (D) STAT6, and (E) negative for SMA. (F) The excision specimen shows a widely excised well-circumscribed nodule.SMA, smooth muscle actin; STAT6, signal transducer and activator of transcription 6.

Solitary fibrous tumour (SFT)

SFTs were first described in the pleura but can arise in any anatomical site, typically in middle-aged adults.2 Only 25 cases of SFT in the breast have been reported. Breast SFT may present clinically as well-demarcated, firm, slow-growing, non-tender lesions. On imaging, SFTs may appear similar to fibroadenomas, but in some cases may be distinguished by the presence of a vascular pedicle and prominent vessels.3

Histologically, SFTs have spindle to ovoid cells with a haphazard or focally storiform arrangement in fibrous or myxoid stroma.1 On immunohistochemistry, the lesional cells are diffusely and strongly positive for CD34 and negative for SMA. SFTs may be distinguished from lesions sharing this immunohistochemical profile (table 1) by their branching ‘staghorn’ vasculature and immunoreactivity for STAT6. An intrachromosomal rearrangement producing a NAB2-STAT6 fusion gene is considered to be a major driver of SFT pathogenesis.4

A minority of SFTs have malignant behaviour. A study of SFTs from across body sites found a 5-year metastasis rate of 34% and a 5-year overall survival rate of 84%.5 Higher incidence of metastasis was correlated with tumour size >10 cm and a mitotic rate >4/10 high-power fields.5 More recently, Demicco et al 6 developed a model predicting the risk of metastases based on tumour size, mitotic rate, patient age and tumour necrosis. Additional histological features proposed to correlate with malignant behaviour include infiltrative margins, hypercellularity and pleomorphism.4 Malignant SFTs may also lose CD34 expression.4 Histological features associated with malignancy were present in only two reported breast SFTs, for which long-term outcomes are unknown.7 8

Metastatic SFTs have also been identified in patients whose primary lesion lacked malignant histological features,9 such that the absence of worrisome features cannot be used to rule out the possibility of metastatic behaviour. This limitation has resulted in even benign-appearing SFTs being considered lesions with uncertain biological potential. Complete excision with clear margins and long-term follow-up are therefore recommended for all SFTs.4 Radiation therapy is generally not indicated for SFTs lacking malignant histological features given their low rates of recurrence, but may be considered for SFTs with malignant histological features.10

Case 2

A 46-year-old premenopausal woman presented with a painful breast mass. Ultrasound demonstrated a hypoechoic, well-circumscribed, mildly lobulated solid mass (figure 2A) that had grown from 3.6 cm to 8.8 cm over a 4-year period. Percutaneous core biopsies of the lesion at two separate times showed features of a bland spindle cell lesion that was favoured to be a fibroadenoma. Because of the large size and ongoing growth of the mass, excision was carried out. The excision specimen contained a 9.2 cm lobulated, tan, rubbery mass (figure 2B). Histology showed fibrotic stroma with slit-like spaces lined by hyperchromatic spindle cells. The spindle cells had no significant atypia or mitotic activity (figure 2C). Unremarkable ducts and lobules were entrapped within the lesion. On immunohistochemistry, the spindle cells were positive for SMA (figure 2D) and CD34 (figure 2E). The patient was diagnosed with nodular PASH and returned to regular screening mammography.

Figure 2

Nodular pseudangiomatous stromal hyperplasia. the lesion appears heterogeneous and mildly lobulated on (A) ultrasound and (B) gross examination. (C) The excised lesion consists of fibrotic stroma with slit-like spaces lined by spindle cells (H&E). On immunohistochemistry, the spindle cells are patchy positive for (D) SMA and diffusely positive for (E) CD34. SMA, smooth muscle actin.

Nodular pseudoangiomatous hyperplasia

PASH is a benign reactive spindle cell lesion most commonly found in premenopausal women, although it can also occur in men with gynecomastia.11 PASH is most often an incidental finding, present microscopically in up to 23% of breast biopsies12 and often omitted from the final diagnosis if more significant pathology is identified. PASH forming a palpable mass (ie, nodular PASH) is more difficult to diagnose on core biopsy, as concerns may be raised about whether the intended target was sampled. Approximately 200 cases of nodular PASH have been reported, with single nodules measuring up to 35 cm.13

Clinical and radiological features associated with PASH may initially be concerning for malignancy. For instance, patients with PASH may present with rapid lesion growth,11 worrisome nipple discharge14 or overlying skin changes resembling peau d’orange.12 Nodular PASH may have ill-defined borders and mixed echogenicity on ultrasound, though more commonly nodular PASH appears well circumscribed and hypoechoic.3 15

Histologically, PASH consists of dense fibrous stroma with slit-like spaces lined by a single discontinuous layer of spindle cells.1 The spaces are pseudovascular and lack red blood cells, unlike the vessels in SFT. The borders of PASH may be histologically ill-defined, similar to the focally infiltrative appearance of NF, reactive spindle cell nodule and neurofibroma. This is in contrast to the well-circumscribed borders of most other benign spindle cell lesions.

On immunohistochemistry, the spindle cells are positive for CD34 and often focally positive for SMA, oestrogen receptor and progesterone receptor.1 14 Hormone receptor expression is consistent with the notion that PASH results from an aberrantly enhanced response of myofibroblasts to hormonal stimuli.11 Immunoreactivity for CD34, SMA and hormone receptors also occurs in myofibroblastoma, a lesion that may be distinguished from PASH by its lack of pseudovascular spaces. Fibroadenoma and phyllodes tumour may have immunoreactivity for SMA and CD34, but have distinctive epithelial components.

PASH occasionally recurs, and is more likely to recur if incompletely excised.11 PASH that is biopsied but not excised tends to remain stable.15 Recently published guidelines from the American Society of Breast Surgeons advise that asymptomatic PASH diagnosed on core biopsy with concordant imaging findings should not be routinely excised.16

Case 3

A 35-year-old woman presented with a rapidly enlarging mobile, firm, oblong breast mass. She had a remote history of bilateral surgical drainage for breast infections. Ultrasound identified a 4.4 cm circumscribed hypoechoic solid mass with small cystic areas and internal vascularity (figure 3A). A core biopsy of the mass revealed spindle cell stroma with mild lymphoplasmacytic inflammation, scattered extravasated erythrocytes and occasional mature adipocytes (figure 3B). Neither significant mitotic activity nor cytologic atypia was present. On immunohistochemistry, the spindle cells were positive for SMA (figure 3C), patchy positive for CD34 (figure 3D) and negative for ER, pancytokeratin, p63 and S100. Adjacent to the spindle cell stroma were areas suspicious for atypical ductal hyperplasia, prompting surgical excision of the lesion.

Figure 3

Nodular fasciitis. (A) On colour Doppler ultrasound, the lesion appears circumscribed and hypoechoic with internal vascularity and cystic areas. (B) On core biopsy, the lesion contains spindle cell stroma with mild lymphoplasmacytic inflammation (H&E). On immunohistochemistry, the spindle cells are (C) positive for SMA and (D) patchy positive for CD34. (E) The excision specimen shows a narrowly excised translucent Tan mass with small cystic areas. SMA, smooth muscle actin.

The excision specimen contained a firm, ill-defined mass with a translucent central tan/grey area surrounded by small cystic structures (figure 3E). Histological features were similar to the findings on core biopsy. Areas of usual ductal hyperplasia were entrapped by the spindle cell lesion, but no residual atypical ductal hyperplasia was identified. Cytogenetic analysis for USP6 rearrangement by dual-colour break-apart probes (Empire Genomics) was negative. However, given the histomorphology of the excision specimen, the lesion was diagnosed as NF. No further clinical intervention was required.

Nodular fasciitis

NF is a benign proliferation of myofibroblasts believed to develop in response to injury, although a history of trauma is often not documented.17 NF most commonly arises in the soft tissue of the trunk, neck and extremities of young and middle-aged adults. There are only 31 reported cases of NF occurring in the breast. Mammary NF presents as a firm painless mobile nodule that rapidly enlarges over days or weeks. The appearance of NF on imaging is variable and can have features concerning for malignancy, such as spiculation on mammography or an echogenic halo on ultrasound.3

Histologically, NF consists of spindle cells in short fascicles or a focally storiform pattern.1 The spindle cells are embedded in a fibrous or myxoid stroma with extravasated erythrocytes and scattered lymphocytes. The margins may be irregular and focally infiltrative with entrapped ducts and lobules at the periphery. The mitotic activity of the spindle cells is variable and may be abundant, contributing to the misdiagnosis of NF as sarcoma.18

On immunohistochemistry, the spindle cells are positive for SMA and typically negative for CD34.1 Weak or patchy positivity for CD34, as in our presented case, may rarely be seen.17 Other lesions that are positive for SMA and negative for CD34 include desmoid-type fibromatosis, reactive spindle cell nodules and inflammatory pseudotumours. The latter two lesions and NF are unique compared with other benign spindle cell lesions as they have a more prominent lymphocytic infiltrate. Among these three lesions, NF may be distinguished by its feathery or tissue culture-like arrangement of spindle cells, areas of myxoid stroma and extravasated erythrocytes. Desmoid-type fibromatosis, in contrast, has distinctive long sweeping fascicles, thin-walled vessels and nuclear immunoreactivity for β-catenin.

Malignant lesions that are typically positive for SMA and negative for CD34 on immunohistochemistry include myofibroblastic sarcoma and fibromatosis-like spindle cell carcinoma (table 1). High-grade forms of these lesions are readily distinguished by the presence of cytologic atypia. Low-grade fibromatosis-like spindle cell carcinoma may be distinguished by its immunoreactivity for p63 and cytokeratin, whereas low-grade myofibroblastic sarcoma may be distinguished by its more homogenous appearance.19 In cases that remain unclear, fluorescence in situ hybridisation may be helpful to detect USP6 rearrangement,20 although 8%–9% of NF cases do not demonstrate a USP6 rearrangement.21 22

Spontaneous resolution of even large NF lesions typically occurs over several weeks, and there have been no reported cases of recurrence.17 19 20 Non-operative management is therefore recommended if the imaging and biopsy findings are consistent with NF.


Benign spindle cell lesions of the breast are histologically similar-appearing entities with diverse underlying biology. Consequently, these lesions require clinical management tailored to the diagnosis. The advancement of ancillary testing, including immunohistochemistry and cytogenetics, may facilitate accurate diagnosis at the time of core biopsy and allow for optimal management. The presented cases demonstrate how detailed histomorphological examination and immunohistochemical profiles can be integrated to distinguish these lesions, and underscore the importance of a multidisciplinary approach to the evolving management of spindle cell lesions of the breast.

Take home messages

  • Benign spindle cell lesions of breast may be distinguished on percutaneous core biopsy sample through careful evaluation of histologic features and immunohistochemical panel of p63, pancytokeratin, SMA and CD34.

  • Use of immunohistochemistry for STAT6 is helpful in distinguishing SFT from other CD34-positive spindle cell lesions.

  • Asymptomatic nodular PASH may be managed with clinical and radiographic follow-up and does not require excision.

  • Use of cytogenetics assessment for USP6 rearrangement may help distinguish nodular fasciitis from other myofibroblastic proliferations.



  • Handling editor Cheok Soon Lee.

  • Contributors JN: collaborated with CS to determine the manuscript structure, compiled case study data, acquired the SFT gross pathology images, compiled and formatted images for the figures, reviewed the literature and drafted the manuscript.CC, MG: recommended inclusion of the selected cases in this study, collected clinical information and revised the content of the manuscript. DC: reviewed radiology images and selected images for inclusion in figures, reviewed pertinent radiology literature and revised the content of the manuscript. CS: conceived of the case series format, diagnosed the PASH and NF patients, acquired histology images and the PASH and NF gross pathology images, suggested the inclusion of table 1 and revised the content of the manuscript.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.