Undifferentiated Pleomorphic Sarcoma Mimicking Breast Implant-Associated Anaplastic Large Cell Lymphoma

Abstract

Breast implant augmentation is a low-risk procedure with few life-threatening complications, most commonly rupture or leakage and contraction of the capsule. Breast implant-associated (BIA) malignancies are rare. Anaplastic large cell lymphoma (ALCL) is the most well-known neoplastic condition associated with breast augmentation. Carcinomas arising in association with implants have been reported but are rarer than ALCL. BIA-mesenchymal tumors are extremely rare and most are locally aggressive fibromatosis. To date, only eight cases of BIA sarcomas have been reported. Herein, we describe a case of silicone BIA-undifferentiated pleomorphic sarcoma (UPS) that was initially mistaken for ALCL because of a significant clinical and radiological overlap in presentation and imaging. Here, we present the morphological and molecular features of this rare neoplasm. We reviewed the existing literature related to BIA sarcomas to highlight the importance of considering this diagnosis in cases of recurrent ALCL-negative BIA effusions.

INTRODUCTION

Since the first procedure using silicone implants was performed in 1962, breast augmentation has evolved using a variety of components and shells [1]. The British Association of Aesthetic Plastic Surgeons’ annual audit revealed that the average annual number of breast augmentation procedures performed in the United Kingdom from 2010 to 2022 was 7,550 [2]. Breast augmentation with silicone or saline implants is known to be a low-risk procedure with few life-threatening complications, most commonly rupture or leakage and contraction of the capsule [3].

Breast implant-associated (BIA) malignancy is extremely rare, with anaplastic large cell lymphoma (ALCL) being the most well-known [4,5]. Carcinomas arising associated with implants are rarer than BIA-ALCL, and have so far been squamous cell carcinomas [6].

BIA-mesenchymal tumors are extremely rare, with only 41 reported cases. Thirty-three of these cases were fibromatosis [7,8,9], and only eight were BIA sarcomas (Table 1) [8,10,11,12,13,14,15]. These include two angiosarcomas, two undifferentiated pleomorphic sarcomas (UPSs), previously termed malignant fibrous histiocytoma/pleomorphic sarcomas, and two fibrosarcomas. In addition, there are dermatofibrosarcoma protuberans with fibrosarcomatous transformation and stromal sarcoma [10,11,12,13,14,15,16]. The latent period for implant-associated fibromatosis is three years [7], but for ALCL and sarcomas, it is at least 10 years [5,8].

Table 1. Reported cases of implant-associated sarcoma.

Publication [reference]	Age	Biomaterial	Time (yr)	Diagnosis
Kobayashi et al., 1988 [10]	55	Silicone injection	19	Malignant fibrous histiocytoma
Wada et al., 1988 [11]	55	Silicone implant	20	Stromal sarcoma
Roncadin et al., 1998 [12]	51	Silicone implant	6	Angiosarcoma
Aharon-Maor et al., 1998 [13]	42	Silicone implant	15	Fibrosarcoma high grade
Cuesta-Mejías et al., 2002 [14]	41	Silicone implant	1.6	Angiosarcoma
Balzer and Weiss, 2009 [8]	59	Silicone implant	1.5	Pleomorphic sarcoma
Balzer and Weiss, 2009 [8]	52	Silicone implant	2.5	Fibrosarcoma
Haghayeghi et al., 2020 [15]	43	Silicone implant	4	Fibrosarcomatous DFSP

Time refers to the interval between breast augmentation procedure and diagnosis.

DFSP = dermatofibrosarcoma protuberans.

UPS is defined by the lack of identifiable line of differentiation despite exhaustive histological and molecular testing, constituting less than 5% of all sarcomas in adults. The incidence of UPS in the breast is extremely rare and not well documented, but several cases have been reported as either primary or secondary to radiotherapy [16,17,18,19,20]. Herein, we describe a case of UPS associated with silicone breast implants, which would bring the total number of BIA-mesenchymal tumors to 42 and the total number of BIA-sarcomas in the literature to 9.

CASE REPORT

A 75-year-old woman underwent bilateral breast silicone implant augmentation 14 years previously for cosmesis. She presented in 2018 with sudden onset painful swelling in right breast, arising on a background of ruptured implant and silicone lymphadenitis. Skin erythema but no subcutaneous nodules, eruptions, erosions, or ulcers were observed. The patient’s left breast was normal. Her previous medical records were unremarkable with no history of breast cancer or radiotherapy. There was no known family history of breast malignancy. Computed tomography (CT) revealed a large peri-implant effusion in the right breast. No mass was detected CT (Figure 1A). Ultrasound (US)-guided fine-needle aspiration (FNA) was performed and ALCL negativity was confirmed. FNA revealed suboptimal, poorly preserved, and paucicellular samples containing blood and nonviable cells. Macrophages and, occasionally, small lymphocytes were also observed. Flow cytometric analysis showed a viability of 49% and within this there were no clonal B cells or evidence of CD2/CD30+ve T cells.

Figure 1. Computed tomography imaging of implant associated sarcoma. (A) There is a large amount of peri-implant effusion with deformed implant shell of right implant. The left implant is intact. (B) After bilateral surgical exploration and excision of the implants bilaterally, heterogeneously enhancing nodular masses are shown in her right breast (red arrow), infiltrating to the chest wall (yellow arrows). The left breast implant had normal appearance during surgery, and the left breast showed post procedural fibrosis only.

An implant-associated infection was clinically suspected, and antibiotic treatment was administered, resulting in temporary relief of symptoms. However, breast pain and swelling recurred, and treatment was reinitiated. The symptoms and peri-implant fluid collection did not resolve, despite 18 months of repeated antibiotic/anti-inflammatory therapy, and a decision was made to surgically excise the implant capsule bilaterally.

Finally, exploration was performed, and the surgical findings revealed an abnormal implant showing marked thickening and adhesion to the breast tissue. Complete surgical excision could not be achieved; this was confirmed on postoperative CT, which showed residual masses within the breast (Figure 1B) and indeterminate nodes in the right axilla. The patient underwent right mastectomy and completion capsulectomy. No evidence of metastasis was observed.

Histological examination of the implant capsule was initially performed, and the subsequent mastectomy showed similar features (Figure 2). The sections revealed nodular masses of spindle cells (ranging from 12 to 21 mm in maximum dimension) arising from the inner aspect of the fibrous implant capsule with extensive fibrino-necrotic material. The nodules were composed of irregular interlacing fascicles of highly atypical and pleomorphic spindle cells with tumor giant cells and brisk mitotic activity (20/10 high-power field) (Figure 2B). Spindle cells infiltrated the full thickness of the implant capsule and invaded the adjacent breast tissue (Figure 2C and D). Small lymphocytes surrounded the tumor and infiltrated the malignant spindle cells with a focal foreign-body-type reaction, indicating silicone rupture. No significant acute inflammation or atypical lymphocyte infiltration was observed.

Figure 2. Undifferentiated pleomorphic sarcoma arising within the breast implant (hematoxylin and eosin stain). (A) Low power view shows sheets of spindle cells with scattered bizarre tumor giant cells forming nodules at the internal aspect of the implant capsule (*). (B) High power view of tumor cells showing the highly pleomorphic nuclei, bizarre tumor giant cells with prominent nucleoli (inset) and frequent mitosis. (C) Tumor cells penetrated the thick, acellular, focally calcified, collagen bundles of the implant capsule (*) and infiltrated the adjacent adipose tissue where there was a florid lymphoctic response. (D) The tumor infiltrates the breast parenchyma and is present between breast glandular/acinar structures (arrows).

Overall, immunohistochemical analysis was not helpful. This revealed a mixed population of B and T lymphocytes that were entirely negative for CD30 and ALK1, thereby excluding the possibility of BIA-ALCL. The epithelial markers MNF116 and AE1/3 were negative in the spindle cells, but highlighted a few acinar structures around the fibrotic capsule. The pleomorphic cells were also negative for Melan A, S100, ERG, PAX8, EBER, SMA, Desmin, CD31, CD34, CD35, ALK-1 and EMA; therefore, no specific line of differentiation was observed. Ki-67 highlighted a proliferative fraction > 20%.

Molecular analysis confirmed the lack of pathogenic variants in the β-catenin (CTNNB1) gene or adenomatous polyposis coli (APC) genes, which are mutated in desmoid-type fibromatosis (CTNNB1 mutation in sporadic cases and APC mutation in tumors arising in the setting of APC germline mutation). The tumor showed a TERT promoter variant C228T mutation (chr5, 1, 295, 228 C > T) of TERT promoter generating a CCGGAA/T or GGAA/T motif, which is a possible transcription factor binding site for Ets transcription factors. A likely pathogenic TP53 variant was also identified (Figure 3). No other structural variants were detected in the DNA-based next-generation sequencing analysis of mutations in any of the genes tested, including BRCA1, BRCA2, CDH1, PALB2, PTEN, STK11, or ATM.

Figure 3. TERT and TP53 mutations in breast implant associated fibrosarcoma, Binary Alignment Map files showing the TERT promoter C228T variant (A) and the TP53 c.503A>T p.(His168Leu) variant (B). Variant allele frequency 20% and 22% respectively.

The immunomorphological features and molecular results were consistent with those of UPS. The tumor was completely excised and staged as pT2N0 G3.

Treatment and follow up

Postoperative US confirmed the absence of residual or recurrent tumor. The patient remained alive and well 43 months after initial presentation, with no evidence of disease recurrence or distant metastasis.

DISCUSSION

Here, we present a case of UPS arising within the capsule of a silicone breast implant in a woman with no history of breast disease or radiotherapy. Her initial presentation mimicked BIA-ALCL, both clinically and radiologically.

The common presentation of BIA-ALCL is breast asymmetry or swelling secondary to late onset seroma, defined as > one year post-implant (85%), or, in rare instances, as a palpable mass—15% (median of 10 years post-implant), or as combination of mass and effusion. This was the same presentation as that in our case, together with pain, erythema, and skin changes. The latter is also found in one-third of patients with ALCL [5,21,22]. Radiologically, the CT appearance of this UPS was comparable to the reported imaging findings of BIA-ALCL, which include peri-implant effusion, contractures, or implant-associated masses [23]. The small soft tissue component was not evident on CT or US examination. In retrospect, breast magnetic resonance imaging, which has the advantages of higher sensitivity and better soft-tissue resolution, may have identified the solid component earlier.

The calculated latency periods for developing BIA-sarcomas and mass-forming BIA-ALCL are similar: 19 months to 20 years, with a median of 10 years [4,8]. In this case, the latency was 14 years.

Although BIA-ALCL can arise in association with both saline and silicone implants [5], all reported sarcomas were associated with silicone implants (Table 1) [8,10,11,12,13,14,15], and accompanied by gel leak/rupture into the surrounding tissue, including two previously reported BIA-UPS cases [8,10]. Furthermore, one patient with fibrosarcoma and one with angiosarcoma had a history of radiotherapy for primary breast carcinoma 7–10 years previously [8,12].

Mutational analysis of BIA-UPS revealed TERT and TP53 mutations. TERT promoter mutations are rare in soft tissue tumors [24], being present most commonly in solitary fibrous tumor and myxoid liposarcoma, and are all C228T mutations, similar to our case, thereby increasing telomerase activation, oncogenesis, and cell immortalization.

TP53 is a tumor suppressor that induces the transcription of genes associated with cell cycle arrest, apoptosis, and metabolism. A putative disruptive event in TP53 was reported in 68% of undifferentiated sarcomas, and one primary breast UPS that underwent molecular analysis also harbored a p53 mutation [18]. This is believed to be an early event that provides a permissive background for widespread genomic loss. The combination of TERT and TP53 mutations was reported in pleomorphic dermal sarcoma, but these also harbor NOTCH1/2 and FAT1 mutations in 50%–70% [25]. Mutations in NOTCH1/2 and FAT1 were not detected in our case.

Balzer and Weiss [8] reported no metastasis in patients with BIA-sarcoma (median follow-up was 3.2 years-range, 16–92 months), but three of seven patients developed recurrences and died of the disease within two years of diagnosis. Four patients were alive with no evidence of disease for a maximum of 30 months during follow-up. Our patient remains disease-free, with no evidence of recurrence or metastasis 43 months after initial presentation.

An association between sarcomas and foreign materials has been shown in animal studies, with one-third of mice developing silicone-induced sarcomas after a latency period of 24 months [26]. Our patient lacked a history of radiotherapy and showed features that conformed to the principles accepted for biomaterial-associated sarcomas, which are a tumor arising in the immediate site of the biomaterial, after an appropriate latency period. However, two studies based on the National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) Program database for the years 1973–1990 revealed that there is no risk of breast sarcoma development in women with silicone breast implants [27,28]. This indicates that there is insufficient evidence to suggest that sarcomas arising in or around the silicone implant capsule are causally related.

Further studies on large (and more recent) databases may provide the long-term evidence needed to ascertain the relationship and mechanism by which rare patients develop BIA-sarcomas.