Extra-thoracic, extra-meningeal solitary fibrous tumours - A case series and service review

Abstract

Background

Solitary fibrous tumours (SFT) are a type of mesenchymal tumour. Whilst the majority of cases follow an indolent course a significant proportion of patients suffer metastases or disease recurrence post-surgical excision. Due to the unpredictable clinical course follow up duration and intensity remains contentious.

Aims

We aimed to determine current outcomes of management of this tumour, apply and assess current risk recurrence models to determine if our standard of care could be improved upon.

Methods and patients

A prospective database of patients treated at a regional musculoskeletal oncology service was assessed. Only extra-pleural, extra-meningeal SFTs were included in the study. Surgical outcome and post-operative investigations were scrutinised and the Pasquali and Demicco recurrence risk models were applied and assessed.

Results

From 2009 to 2019 12 patients were identified, 8 female and 4 males. Their age at diagnosis ranged from 21 to 76 years. 11 patients underwent surgery with curative intent and no patient suffered disease progression or recurrence, with a mean follow up time of 41 months. One patient presented with metastatic disease and was managed palliatively.

Conclusions

Following this review of our case series and utilising risk recurrence models published in the literature we have changed our follow up protocol. In new cases of SFT the Pasquali prognostic model, with the addition of the presence or absence of necrosis, will be utilised. If a patient has benign features on initial biopsy we propose to not perform staging. Furthermore, if biopsy and final pathology results remain concordant, with no concerning features, and the patient has undergone complete excision reduced intensity follow up could be considered.Level of evidence Level IV, retrospective case series.

1. Introduction

Solitary fibrous tumours (SFT) are a type of mesenchymal tumour. They were previously thought to occur predominantly within the thorax, but are now known to occur as frequently in extra-thoracic locations.1, 2, 3 Previously SFTs were termed haemangiopericytoma, until 2002 when the third WHO classification of soft tissue tumours grouped extra-meningeal SFTs, haemangiopericytoma, lipomatous haemangiopericytoma and giant cell angiofibroma within the spectrum of extrapleural SFT.⁴ The recurrent gene fusion NAB2-STAT6 on chromosome 12q13 was identified from whole exome sequencing data as a unique mutation to SFTs.^5,6

SFTs arise from spindle cells and histopathologically range from low grade to malignant and high grade.⁷ Accordingly, the majority of the tumours follow a benign course. However, a significant proportion of patients suffer disease progression, with distant metastases described to many sites inclusive of lung, liver and bone.⁸ A recent multi-centre study of extra-thoracic and extra-meningeal SFTs reported an 18.5% recurrence rate, with the majority of these recurrences occurring distally from the primary tumour site.⁹

Traditional sarcoma staging systems, such as the American Joint Committee on Cancer staging, have not been validated for predicting the behaviour of SFTs.¹⁰ The majority of SFTs follow an indolent course, however, a small but significant proportion of cases demonstrate overt sarcomatous transformation. Therefore, the behaviour of SFTs is commonly described as unpredictable.¹¹ The main modality of treatment for SFTs is surgical excision, aiming to achieve complete resection with negative margins. The use of adjuvant therapy is not commonly utilised with conventional chemotherapy regimens, such as anthracyclines, shown to not improve outcomes.¹² Chemotherapy and radiotherapy are more often used in metastatic disease or in cases with positive margins post excision,.^7,9 A recent large retrospective, international study of SFTs of all extrameningeal locations demonstrated improved local disease control, without survival benefit, with the use of adjuvant radiotherapy.¹³ Marginal excision is often reported, this has been attributed to SFTs commonly being deep seated with the presence of large collateral feeding vessels.⁶ Metastatic disease has also been reported to occur several years post successful surgery.^8,14,15 Due to the above factors follow up post-surgery often advocates reviews over a long period of time, inclusive of local examination and chest surveillance imaging.

It is evident that the clinical course of patients suffering from SFTs is highly unpredictable and follow up duration and intensity remains contentious. Several studies have proposed risk recurrence models, in order to stratify intensity and duration of follow-up for patients suffering this disease.^9,16 We performed a review of all cases of SFT within our tertiary referral musculoskeletal oncology centre. We aimed to determine current outcomes of management of this tumour, apply and assess current risk recurrence models and determine if our standard of care could be improved upon.

2. Methods

A prospective database of patients treated at a regional musculoskeletal oncology service was assessed. All patients managed within the service with a diagnosis of SFT were identified between the years of 2010 and 2019. Only patients with a histological diagnosis, from biopsy and excision, of SFT were included. Only extra-pleural, extra-meningeal SFTs were included in the study. The tumour size and location and extent of preoperative staging was assessed. Tumour characteristics including presence of necrosis, cellularity, mitotic rate and degree of pleomorphism were assessed for each patient. The type of surgical excision, and tumour margins achieved was assessed along with the use of any adjuvant therapies. Post-operative records and investigations were scrutinised to determine outcomes of management of the disease for each patient.

To determine whether the services post-operative management could be improved upon two recurrence risk models for SFT were applied to the cohort.

3. Results

12 patients were treated for SFT within our centre between 2009 and 2019. 8 were female and the mean age at diagnosis was 53 years (range 21–76 years). Locations and tumour indices are included in Table 3, Table 4. All patients underwent image guided biopsy. Mean tumour size was 7.7 cm (range 2.2–16 cm). 9 patients underwent staging with computerised tomography (CT) scanning of chest, abdomen and pelvis. No discrepancy was observed between the pre-operative biopsy diagnosis and post-operative tumour specimen. Fig. 1 demonstrates pre-operative magnetic resonance imaging (MRI) of a patient within the series.

Table 3.

Anatomical location of SFTs within series.

Location	Number of cases	%
Upper Limb	2	16.7%
Lower Limb	5	41.7%
Torso	1	8.3%
Pelvic	4	33.3%

Table 4.

Tumour indices of SFTs within series.

	N (%)
Necrosis present	2 (16.7%)
Mitotic rate
Low <4 HPF	10 (83.3%)
High >4 HPF	2 (16.7%)
Cellularity
Low	2 (16.7%)
Moderate	1 (8.3%)
High	9 (75%)
Pleomorphism
Low	10 (83.3%)
Moderate	1 (8.3%)
High	1 (8.3%)

Fig. 1.

Coronal and axial pre-operative MRI scan of a patient with a benign SFT of the thigh.

One patient presented with distant bony and pulmonary metastases at time of diagnosis and was managed palliatively, their biopsy confirmed a malignant SFT. The remaining 11 patients were treated with curative intent with complete tumour excision.

No neoadjuvant or adjuvant therapy was utilised in addition to tumour resection. One patient presented with distal metastases. She underwent palliative chemotherapy followed by prophylactic intra-medullary nailing of a femur and open reduction and internal fixation of a humerus for metastatic bone disease. This patient died 18 months after diagnosis.

The overall survival of the cohort is a median of 58 months (range of 11–141 months). Of the 11 patients managed with curative intent complete excision was achieved in all the primary operations. No local recurrences or distant metastases have occurred. 6 patients underwent chest surveillance at follow up via chest radiographs. The mean follow-up post operation was 41 months. 4 patients remain under active follow up.

Pasquali et al.’s SFT recurrence risk model, Table 1, was applied.⁹ Recurrence risk was categorized as very low, low, intermediate and high in 2 (16.7%), 7 (58.3%), 2 (16.7%) and 1 (8.3%) respectively. The patient who presented with metastatic disease was identified as high risk.

Table 1.

Pasquali et al. recurrence risk model. Tumour indices and risk of recurrence displayed.¹⁰

Variables		Score
Mitotic rate	Low High (>4 x HPF)	0 3
Cellularity	Low Moderate - High	0 2
Pleomorphism	Low Moderate - High	0 2

Recurrence Risk	Total score
Very low	0
Low	2
Intermediate	3–5
High	>5

Subsequently the recurrence risk stratification scheme by Demicco et al., Table 2, was applied.¹⁶ Risk was categorized as low in 7 patients (58.3%) and intermediate in 5 (41.6%) with no patient scoring within the high risk group.

Table 2.

Demicco et al. risk stratification model. Patient and tumour indices measured and recurrence risk classification displayed.²¹

Risk Factor	Score
Age
<55	0
>55	1
Tumour size (cm)
<5	0
5 to < 10	1
10 to <15	2
>15	3
Mitotic rate (/10 HPF)
0	0
1–3	1
>/ = 4	2
Tumour necrosis
<10%	0
>/ = 10%	1
Risk class	Total score
Low	0–3
Intermediate	4–5
High	6–7

4. Discussion

In comparison to many previously published series, only extra-thoracic and extra-meningeal SFTs were included, potentially improving the relativity of these results to musculoskeletal oncologists. Importantly all diagnoses were made after 2002, and thus all tumours meet the current SFT criteria.⁴ Furthermore, all patients were managed at one hospital excluding variability between surgeons of different centres, which may have occurred within larger, multicentre series.^3,9 A significant proportion of patients in previous series have had re-excision of the tumour due to unknown or positive margins.^3,17 No patient in this series received neo-adjuvant or adjuvant therapy in combination with surgery with curative intent. However, this study has a relatively small sample size in comparison to other series, with multi-site series having in excess of 240 patients.⁹ Table 5 demonstrates a comparison of this series to other single site series.

Table 5.

Summary of single site SFT series published in the literature. Outlining the number of patients, malignant status, follow up and survival data where available. NA = not available.

Study	Number of patients	Primary tumour location	Benign/Malignant	Median Follow- up (months)	5 year overall survival
Spitz FR 199818	36	All sites	NA	57	71%
DeVito 201519	82	Extrameningeal	42/40	55	65%
O'Neill 201720	123	All sites	59/64	124	NA
Cranshaw 200921	33	Extrathoracic	15/18	43	66%
This series	12	Extrathoracic, extrameningeal	11/1	41	NA

The heterogeneity of the series in Table 5 does make comparisons difficult. However, extrathoracic SFT disease is often noted to have more aggressive features on pathological assessment and worse disease free survival.²⁰ This is important for our series, which only has assessed the outcomes of extrameningeal and extrathoracic SFTs.

All tumours were excised completely on primary procedure and no recurrences were encountered. The length of follow up was comparable to other studies, where a 18.5% recurrence rate was observed.^9,17 Distant disease progression has been described in patients with SFT, in this series only 50% of patients underwent chest surveillance with radiographs.

The outcome of excision of SFTs has been reviewed in several series with confliction over prognostic factors. Tumour size >10 cm and high mitotic rate were associated with a significant increase in metastatic rate in Van Houdt et al.’s series.³ However, this series included intra-thoracic SFTs. Conversely Pasquali et al. found larger tumour size to be associated with a better prognosis and reduced recurrence rate. Furthermore, Pasquali et al.’s series showed on multivariable analysis that a high mitotic rate, cellular atypia and hypercellularity were significantly associated with disease recurrence.⁹ These key variables were used in their recurrence risk model, Table 1.

A further risk stratification scheme was proposed by Demicco et al., in 2012.¹⁷ This suggested that small tumour size and low mitotic rates are highly unlikely to metastasize, whereas large (>15 cm) tumours in patients >55 years of age with mitotic figures ≥4/10 high-power fields have a high risk of both metastasis and death, and require close follow up. This group then verified and refined their model, Table 2, and found that adding tumour necrosis improved the power of their model.¹⁶ Whilst this cohort includes intra-abdominal and intra-thoracic SFTs it further suggests that follow up can be stratified by tumour characteristics.

Using their risk recurrence model, Pasquali et al. advised upon reduced follow up scheduling and potentially non-operative management for the treatment of specific cases of SFT. Patients with low mitotic rate, cellularity and pleomorphism appeared at virtually no risk of recurrence, and thus low intensity follow up or non-operative management of unfit patients could be recommended.⁹ Contrastingly, high mitotic rate, cellularity and pleomorphism were features of high risk of recurrence necessitating longer term follow up. Applying this several patients within our dataset could have been discharged post-operatively or if deemed a high anaesthetic risk offered non operative management. The Pasquali recurrence model was not validated by the authors in a separate series, a noted weakness of their series, but on application to our small case series seems to be appropriate for guiding post-operative follow up reviews. Surprisingly Demicco et al.’s recurrence risk model did not identify any patients within this dataset as high risk.¹⁶

Given that no tumour recurrence locally or distant was observed, it could be debated that surveillance with chest radiographs was over used. 6 patients received chest radiographs on follow up review, however only 2 patients were identified as high risk upon using both the Pasquali and Demicco recurrence risk models.^9,16

The presence of necrosis within a soft tissue sarcoma is an important predictor of disease aggressiveness, and it is utilised in the Trojani grading system.¹⁵ However, in SFTs the presence of necrosis and its prognostic significance is contentious.^3,22 Necrosis was not seen to be predictive of disease progression in the largest known case series, with the caveat that a significant proportion of patients had received neo-adjuvant therapy, which may have induced tumour modification.⁹ The presence of necrosis has been included in Demicco et al.’s risk recurrence model, however their series includes SFTs of all anatomical locations.¹⁶ In a large case series of 92 patients with extrathoracic SFTs, by Vallat-Decouvelaere et al., necrosis was associated with, but not predictive, of aggressive clinical behaviour.⁸ Due to the persistence of uncertainty over the importance of the presence of significant necrosis within SFTs, it appears prudent to continue to include it as a potential marker of poor clinical outcome.

The recurrent gene fusion NAB2-STAT6 is believed to be the driver of the disease, and importantly, subgroups of different NAB2-STAT6 fusion types have been recognised. These subgroups have differing prognoses, with the NAB2ex6-STAT6ex16/17 occurring in deep soft tissues of younger patients and displaying more aggressive behaviour and NAB2ex4-STAT6ex2/3 corresponding to the more benign pleural based disease.²³ An important future step in the management of this disease will be when genetic investigations are integrated in to clinical diagnosis and management.

Surgical excision remains the primary treatment modality. Due to the vascular nature of these tumours the significance of antiangiogenic therapies has been investigated. Drugs such as sorafenib and sunitinib, have been used, and shown to be active in patients with locally advanced or metastatic disease.²⁴ Novel targeted therapies to growth signalling pathways, such as Pazopanib have been trialled in patients with metastatic disease, with similar overall survival results noted in comparison to previously reported series.²⁵ The NAB2-STAT6 fusion leads to alteration in growth and development and an overexpression of growth factors, receptor tyrosine kinases and histone deacetylases. It is hoped that these pathways may be therapeutic targets for future therapies. Thermoablation techniques for the treatment of extremity SFTs, such as microwave ablation and high intensity focused ultrasound, have yet to be reported in the literature, however a recent case study by Fiore et al. demonstrated effective treatment of a recurrent pleural SFT via microwave ablation, suggestive that this may be an increasingly investigated technique.²⁶

Following this review of our case series and utilising risk recurrence models published in the literature we have changed our follow up protocol. For new patients we will apply the Pasquali model, but also add the presence or absence of necrosis as an additional risk factor to consider.⁹ Follow up reviews with chest radiographs will be reserved for patients with SFTs which had malignant features on pathological assessment, or who are deemed high risk utilising the Pasquali recurrence model.

Key learning points

•Complete surgical excision remains the mainstay of treatment •Adjuvant therapy remains important for incomplete or unresectable tumours and metastatic disease •Systemic targeted therapies are required •Management of SFTs may become stratified to NAB2-STAT6 fusion types

Declaration of competing interest

Each author certifies that he or she has no commercial associations (e.g. consultancies, stock ownership, equity interest, patent/licensing arrangements) that might pose a conflict of interest in connection with the submitted article. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institution and/or national research committee and with the 1964 Helsinki declaration and it's later amendments or comparable ethical standards. This research was performed at Glasgow Royal Infirmary and the Queen Elizabeth University Hospital.