Keywords
thoracoscopy/VATS - surgery - complications - pleural disease (including drainage)
- chest
Introduction
Solitary fibrous tumors of the pleura (SFTPs) are primary pleural tumors originating
from the mesenchymal tissue. They were firstly described by Wagner et al as primary
neoplasms of the pleura in 1870[1]; then, in 1931 Klemperer and Coleman distinguished localized from diffuse mesothelioma.[2] According to Stout and Murray in 1943, localized mesothelioma had a mesothelial
origin.[3] Later, the development of immunohistochemistry analysis showed that SFTPs arise
from the mesenchymal cell layer and were different from pleural mesothelioma.[4]
Surgical treatment was the first choice for management of SFTPs. However, some tumors
recurred after several years from an R0 resection, but there were no defined guidelines
for the follow-up of these tumors and the postoperative therapy due to the rarity
of these tumors.
In this study, we aimed to analyze the behavior of SFTPs and the predictive prognostic
factors in order to define the follow-up and potential postoperative treatments.
Materials and Methods
This was a retrospective, multicenter study from two high-volume centers in Italy.
Data of patients diagnosed with pleural solitary fibrous tumors (SFTs) between January
2003 and October 2022 were prospectively recorded and retrospectively analyzed. All
patients undergoing R0 surgical resection of pleural SFT were eligible. The aim of
this study was to identify the predictive prognostic factors and the correlation between
tumor characteristics and recurrence.
The exclusion criteria were the following: (1) no R0 resection of initial tumor; (2)
metastatic disease; (3); mesothelioma, lung cancer, or other known cancers; (4) medical
inoperability because of comorbidities; and (5) incomplete follow-up.
Patients were divided in two groups: benign and malignant. According to England et
al's criteria,[5] they were classified as malignant if one or more of these criteria were present:
>4 mitosis/10 high power fields (HPF), necrosis, pleomorphism, high cellularity, and
atypical nuclei.
The difference between two study groups were assessed in order to define whether the
different histology could affect outcomes. Additionally, a univariate and a multivariate
analysis were performed in order to identify survival prognostic factors. The study
was approved by the Local Ethics Committees of the coordinating center of the study.
All the patients gave a written informed consent for the surgical treatment and for
the anonymous use of their data for scientific purpose only.
Study Population
The patients' clinical records were analyzed for age, sex, tumor size, signs and symptoms,
side, positron emission tomography (PET) maximum standardized uptake values (maxSUV),
surgical procedures, postoperative course, pathological features, recurrence, follow-up
period, and survival ([Table 1]). All the patients underwent thoracic computed tomography (CT) scan followed by
flourine-18 fluorodeoxyglucose PET/CT (18F-FDG-PET/CT), and the histological diagnosis was both preoperative, with fine needle
aspiration biopsy, and postoperative. SFTPs were positive for vimentin, CD34, CD99,
and Bcl-2, and negative for keratin.[6]
Table 1
Study population with SFTPs
|
Variables
|
Total population (n = 107), % or IQR
|
Benign (n = 48)
|
Malignant (n = 59)
|
p-value
|
|
Age at diagnosis (y)
|
60 (36–87)
|
58
|
66
|
0.51
|
|
Men
|
58 (54%)
|
28
|
30
|
0.43
|
|
Women
|
49 (46%)
|
20
|
29
|
0.43
|
|
Smoker
|
26 (24%)
|
10
|
16
|
0.45
|
|
Asymptomatic
|
52 (48%)
|
22
|
30
|
0.60
|
|
Symptoms[a]
|
55 (52%)
|
26
|
29
|
0.60
|
|
Chest discomfort
|
34 (32%)
|
12
|
22
|
0.02
|
|
Cough
|
14 (14%)
|
9
|
5
|
0.30
|
|
Dyspnea
|
7 (6%)
|
5
|
2
|
0.17
|
|
Size
|
13 (1–32)
|
13
|
16
|
|
|
>10
|
36
|
20
|
16
|
0.11
|
|
≤10
|
71
|
28
|
43
|
0.11
|
|
Preoperative diagnosis
|
|
maxSUV (range)
|
2.9 (1.4–4.6)
|
1.7
|
3.1
|
0.003
|
|
FNAB
|
64 (60%)
|
25
|
39
|
0.14
|
|
Surgery
|
|
Thoracotomy
|
82 (77%)
|
39
|
43
|
0.30
|
|
VATS
|
25 (23%)
|
9
|
16
|
0.30
|
|
Tumor excision
|
77 (72%)
|
31
|
46
|
0.12
|
|
Wedge
|
26 (24%)
|
15
|
11
|
0.13
|
|
Lobectomy
|
4 (4%)
|
3
|
1
|
0.21
|
|
Pleural location
|
|
Parietal
|
68 (64%)
|
29
|
39
|
0.54
|
|
Visceral
|
39 (36%)
|
19
|
20
|
0.54
|
|
Recurrence
|
20 (19%)
|
4
|
16
|
0.013
|
Abbreviation: FNAB, fine-needle aspiration biopsy; IQR, interquartile range; maxSUV,
maximum standardized uptake values; SFTPs, solitary fibrous tumors of the pleura;
VATS, video-assisted thoracic surgery.
a Some patients presented more than one symptom.
Survival and Recurrence
In-hospital mortality and operative mortality were defined as the number of patients
who died in the first 30 days after surgery. Follow-up consisted of a total body CT
scan every 6 months for the first year and every year for the following years. Overall
survival (OS) was calculated from the time of the resection to the last follow-up
or death. Disease-free survival (DFS) was defined as the time from first tumor excision
to date of recurrence. The site, time, number, and management of recurrences were
also recorded for each patient.
Statistical Analysis
Categorical variables and non-normally distributed variables are presented as number
and percentage and as median and interquartile range [IQR] respectively. χ
2 test and Wilcoxon or Kruskal–Wallis tests were used for statistical analysis. The
Kaplan–Meier method was used to obtain the survival curve. A p-value of less than 0.05 was considered significant. MedCalc statistical software
was used for analysis.
Results
There were 137 patients diagnosed with SFTPs from October 2022 to January 2003. Patients
excluded from the study were those with the following: no R0 resection of initial
tumor (n = 7); metastatic disease (n = 2); mesothelioma, lung cancer, or other known cancers (n = 5); medical inoperability because of comorbidities (n = 7); and incomplete follow-up (n = 9). Finally, 107 patients, comprising 58 men and 49 women, were included in the
study, with a mean age of 60 years (range: 36–87 years). Only 26 of them were smokers.
Fifty-two patients were asymptomatic (48.5%) and 55 had clinical manifestations: the
most common was chest discomfort in 34 patients (31%), 7 patients complained dyspnea
(6.5%), and 14 patients had cough (13%). We noted a correlation between size and clinical
presentation, since most of the asymptomatic patients had tumor size ≤10 cm (71 cases)
and most of the symptomatic ones had tumor size greater than 10 cm (36 cases). The
tumor was detected with thoracic CT scan and then all patients underwent a 18F-FDG-PET/CT. They appeared as rounded well-defined masses. Tumor size was between
1 and 32 cm, in particular, 71 cases between 1 and 10 cm (67%), 26 cases between 11
and 20 cm (24%), 9 cases between 21 and 30 cm (8%), and 1 case greater than 30 cm
(1%). The median size was 13 cm and there was no statistically significant difference
between benign and malignant tumors in terms of size. Its localization was right hemithorax
in 66 cases and left hemithorax in 41 cases. The mean maxSUV was 2.9 for the entire
study population. It was 1.7 (range: 1.4–3.5) for the benign group and 3.1 (range:
2.9–4.6) for the malignant group. Regarding recurrent SFTs, the mean maxSUV was 2.4,
with a range of 1.2 to 3.9. In particular, the mean maxSUV for benign recurrences
was 1.5 and that for malignant recurrences was 2.8 ([Table 2]).
Table 2
Patients with recurrence
|
Variable
|
No. or mean +- SD (% or IQR)
|
Benign (n = 4)
|
Malignant (n = 16)
|
p-value
|
|
Patients
|
20
|
4
|
16
|
0.013
|
|
Disease-free survival (mo)
|
143.3 ± 6.1 (25–175)
|
159 ± 2.9
|
132 ± 8.2
|
0.003
|
|
maxSUV (range)
|
2.4 ± 1.5 (1.2–3.9)
|
1.5 ± 1.3
|
2.8 ± 1.4
|
<0.001
|
|
Pattern of recurrence
|
|
Local
|
7 (35%)
|
3
|
4
|
0.06
|
|
Regional
|
13 (65%)
|
1
|
12
|
0.06
|
|
Distant (extrathoracic)
|
0
|
|
|
|
|
Pleural location
|
|
Parietal
|
12 (60%)
|
2
|
10
|
0.53
|
|
Visceral
|
8 (40%)
|
2
|
6
|
0.54
|
|
Treatment of recurrence
|
|
Excision + wedge
|
15 (75%)
|
2
|
13
|
0.19
|
|
Excision + lobectomy
|
3 (15%)
|
1
|
2
|
0.53
|
|
No surgery
|
2 (10%)
|
1
|
1
|
0.26
|
|
VATS
|
7
|
2
|
5
|
0.48
|
|
Thoracotomy
|
13
|
2
|
11
|
0.48
|
Abbreviation: IQR, interquartile range; SD, standard deviation; VATS, video-assisted
thoracic surgery.
There was a statistically significant difference between maxSUV of benign and malignant
SFTPs (p = 0.003).
Surgical Findings
All the patients were treated with surgery with the aim to obtain R0 resection. We
considered adequate margins of at least 2 cm. An isolated mass excision was performed
in 77 cases (72%), and lung resection was performed in the remaining 30 cases as well.
In 26 cases (24%), mass excision was associated with pulmonary wedge resection, while
in 4 cases (5%), it was associated with pulmonary lobectomy. Lung resection was necessary
when the tumor adhered strongly to the lung parenchyma or infiltrated it. The surgical
approach was thoracotomy in 82 cases and video-assisted thoracic surgery (VATS) in
25 cases. VATS was preferred for small tumors (tumors excised in VATS were <10 cm)
not infiltrating the surrounding structures. Chest tube was removed after 1 to 22
days (mean: 6 days).
Ten patients presented the following postoperative complications: prolonged air leak
(n = 7); anemia for bleeding (n = 2) that did not require a surgical revision; and lobe atelectasis due to retention
of bronchial secretions (n =1).
Pathological Findings
Patients with tumor size between 1 and 5 cm had surgical resection and the histological
diagnosis was confirmed by postoperative examination. In 64 cases (size > 5 cm), the
histological diagnosis was made by CT-guided biopsy. In most cases, they were characterized
by a proliferation of spindle cells with connective tissue. The immunohistochemistry
confirmed the positivity for vimentin, CD43, CD99, and Bcl-2. Forty-eight cases had
malignant features, in particular mitosis >4/10 HPF. In 39 cases, tumors originated
from the visceral pleura and in 68 cases it originated from the parietal pleura.
Recurrence
Twenty of the 107 patients had tumor recurrence. Recurrence was more common among
patients with malignant tumor than among those with benign tumors. Characteristics
of patients with recurrence are summarized in [Table 2]. At a multivariate analysis, histological characteristics (high mitotic index) and
maxSUV were related to recurrence (p = 0.001 and 0.004). There was no statistically significative difference between tumor
size and recurrence (p = 0.72; [Table 3]).
Table 3
Multivariate analysis for recurrence
|
Variable
|
HR
|
95% CI
|
p-value
|
|
Age (>70 y)
|
1.1
|
0.7–2.8
|
0.25
|
|
Comorbidity (yes vs. not)
|
0.8
|
0.4–1.2
|
0.52
|
|
>4 mitosis/10 high power fields (HPF)
|
5.8
|
2.7–4.3
|
0.001
|
|
maxSUV
|
3.7
|
1.8–5.8
|
0.004
|
|
Parietal vs. visceral pleura
|
0.53
|
0.8–2.4
|
0.85
|
|
Tumor size (>10 cm)
|
1.08
|
0.7–1.6
|
0.72
|
Abbreviation: CI, confidence interval; HR, hazard ratio; maxSUV, maximum standardized
uptake values.
The mean DFS was 143.3 ± 6.1 months. DFS of benign and malignant tumors was 159 ± 2.9
and 132 ± 8.2 months, respectively (p = 0.003; [Fig.1]).
Fig. 1 Disease-free survival according to histological features (benign vs. malignant).
Recurrence was more common among patients with malignant tumors than among those with
benign tumors.
We considered the presentation of an SFT after more than 5 years from surgery as a
recurrence, because in all cases the tumor presented the same histologic features
of the first tumor.
In the literature, the definition of recurrence and new onset of a second primary
SFT is not defined in terms of DFS. Furthermore, other authors described cases of
SFTs that recurred more than 5 years after the first tumor.[7]
All recurrence had the same histologic features of the primary tumor including the
benign SFTs, which maintained a mitotic rate inferior to 4 mitosis/10 HPF; 7 were
local and 13 were regional. Two of the patients who presented recurrence did not undergo
surgery because of their poor preoperative status, while 18 of them underwent a second
surgery to excise the tumor. In all cases, tumor excision was associated with pulmonary
resection (15 wedges and 3 lobectomies; [Table 2]). A parenchymal resection was needed because of the tenacious adhesions of the tumor
to the lung and because of the infiltration of the parenchyma, in order to obtain
free margins. In all cases, we obtained clear margins. It could be possible that in
some cases we had an R1 resection that was not recognized at the histological examination,
explaining the recurrence of benign SFT tumors. No death or major complications occurred.
Survival
The median follow-up time was 85 months. OS was 150.9 ± 5.3 months: OS of patients
with benign SFTPs was 143 ± 6.9 months and that of patients with malignant SFTPs was
of 150 ± 6.5 months (p = 0.78; [Fig. 2]). The overall 5- and 10-year survival rates were 95 and 70%, respectively, with
a median survival of 88 months. Only one of them with malign SFTP died of recurrence;
the others died for reasons not related to the tumor.
Fig. 2 Overall survival according to histological features (benign vs. malignant). There
is no significant difference in terms of overall survival.
Three patients died within the first 30 days after surgery for hemodynamic failure.
At a multivariate analysis, the presence of comorbidities (such as cardiovascular
disease, respiratory disease, and other neoplastic disease) was the only significant
prognostic factor (p < 0.001). According to histological features, there were no significant difference
in terms of OS ([Table 4]). Patients with benign SFT can have a shorter survival than patients with malignant
SFTs, because OS is related to the comorbidities (p < 0.001).
Table 4
Multivariate analysis for overall survival
|
Variable
|
HR
|
95% CI
|
p-value
|
|
Age (>70 y)
|
1.5
|
0.1–3.1
|
0.30
|
|
Comorbidity (yes vs. not)
|
3.8
|
2.50–3.75
|
<0.001
|
|
Tumor (benign vs. malignant)
|
0.2
|
0.1–2.1
|
0.43
|
|
maxSUV
|
1.1
|
0.3–3.4
|
0.71
|
|
Parietal vs. visceral pleura
|
0.8
|
0.2–2.1
|
0.35
|
|
Reintervention for recurrence
|
0.34
|
0.7–1.5
|
0.53
|
|
Recurrence
|
0.54
|
0.4–2.5
|
0.74
|
Abbreviation: CI, confidence interval; HR, hazard ratio; maxSUV, maximum standardized
uptake values.
Discussion
Surgical resection is the gold standard management of SFTPs. They require a microscopically
radical (R0) resection. In fact, the literature reports that OS of patients who received
surgery is significantly better. In our study, all patients were treated with an R0
resection and the OS is long both in patients who presented and did not present recurrence.[8] However, a system to predict the risk of recurrence has not been assessed yet. In
the literature, there are only case series or anecdotal reports describing the approaches
of different centers to this neoplasm. We divided the patients into two groups, benign
and malignant, in order to analyze differences in terms of prognosis. The aim of our
study was to analyze the behavior of SFTPs and to predict their recurrence in the
postoperative years based on radiological and histological features, in order to detect
and treat recurrence early.
First, the clinical data of our cases were in line with the previous literature. The mean
age was 60 years, similar to the data in the literature. Most of the patients were
asymptomatic, while symptomatic patients presented nonspecific symptoms such as chest
discomfort, cough, and dyspnea. Symptoms were mainly related to the dimension of the
mass. Thirty-one of the symptomatic patients presented tumors with of a diameter greater
than 10 cm. SFTPs are described to be related with paraneoplastic syndromes such as
Doege–Potter syndrome and a nonislet cell hypoglycemia due to tumor production of
insulinlike growth factor-2 (IGF-2).[9] Also, Pierre–Marie–Bamberger syndrome can be present, due to the secretion of hepatocyte
growth factor (HGF) or hyaluronic acid by tumor.[10]
[11]
[12] None of our patients presented paraneoplastic syndromes.
All patients discovered the presence of an intrathoracic mass through radiological
examination (CT scan) done for other reasons or to investigate the symptoms they complained
about. On CT scan, tumors were homogeneous and defined masses adjacent to the chest
wall, forming with it an obtuse angle ([Fig.3]). In four cases, the tumor was localized in the anterior mediastinum, going in differential
diagnosis with the most frequent mediastinal tumors such as lymphoma or thymoma. In
seven cases, the tumor was in the fissure space mimicking a pulmonary tumor. Larger
masses can mimic other tumors such as sarcomas or big peripheral lung cancer. In our
population, the size of the tumor was between 1 and 32 cm, and there was no relationship
between dimensions and malignancy. CT scan alone is not enough for the diagnosis of
SFTP. All patients who were referred to our attention underwent 18F-FDG-PET/CT. In the literature, there is no definition of standard uptake values
useful to differentiate benign from malign tumors.[13]
[14] In our study, malignant tumors had higher maxSUVs than benign tumors. They were
between 1.4 and 4.6, and in particular maxSUVs of benign tumors were between 1.4 and
3.5 and maxSUVs of malign tumors were between 2.9 and 4.6.
Fig. 3 Radiological presentation of solitary fibrous tumors of the pleura.
Histopathology and immunohistochemistry are necessary for the diagnosis of SFTPs:
they are positive for vimentin, CD34, CD99, and Bcl-2, and negative for keratin[15] and nuclear localization of STAT-6.[16] Tumors arose from the visceral pleura in 39 patients and from the parietal pleural
in 68 patients.
We used the England et al criteria[5] for distinction between benign and malignant tumors and distinguished 48 malignant
tumors from 59 benign tumors, in accordance with some studies in which benign tumors
were more common than malignant tumors.[17]
[18]
[19]
The treatment of SFTPs is surgical, with the aim to obtain a complete resection of
the tumor with negative oncologic margins. Surgical approach varies from case to case.
When the characteristics of the tumor made it possible, we preferred a minimally invasive
surgical approach. We used VATS for tumors of size less than 10 cm, not infiltrating
surrounding structures. In the remaining cases, we preferred a lateral thoracotomy.
In the cases in which the tumor adheres tenaciously to the lung parenchyma, a lung
resection is necessary. The aim is to save as much lung as possible, so wedge resection
is preferred. Anyway, if the tumor invades most of the lobe or if it is central, an
anatomical lobectomy is necessary. In 30 cases, excision of tumor was performed together
with a pulmonary resection, wedge in 26 cases and lobectomy in 4 cases. Surgical outcome
was good since no patient died during surgery and postoperative complications presented
in few cases.
Second, the behavior of SFTPs is hard to predict, since recurrence is not exclusive of malignant
tumors. Different scoring systems have been proposed to predict the risk of recurrence
and adverse outcome. De Perrot staging of SFTPs identifies the following[20]: (1) stage 0, pedunculated tumor without signs of malignancy; (2) stage I, sessile
or inverted tumor without signs of malignancy; (3) stage II, pedunculated tumor with
histologic signs on malignancy; (4) stage III, sessile or inverted tumor with histologic
signs of malignancy; and (5) stage IV, multiple synchronous metastatic tumors. Tapias
et al[21] proposed a scoring system based on clinical and histological features in order to
predict recurrence after SFTP resection. They affirmed that the risk of recurrence
is higher in lesions originating from the parietal pleura. In the meta-analysis conducted
by Liu et al,[22] the recurrence rate for benign SFTP was 3% and that for malignant SFTPs was 22%,
and there was no statistically significant difference between tumors originating from
parietal or visceral pleura. Diebold et al[23] affirmed that a high proliferation rate by MIB-1 (>10%) is associated with adverse
outcome. They also proposed a score based on mitosis, necrosis, tumor size, and MIB-1.
Validation of these scores is necessary in future studies with a larger number of
cases and a longer follow-up time.
An SFT is characterized by a pathognomonic gene fusion between NAB2 (NGFI-A binding
protein 2) and STAT6 (signal transducer and activator of transcription 6), which is
suggested to be the central molecular alteration driving tumor development and progression.
The presence of fusion has been confirmed in nearly all SFTs and the introduction
of STAT6 as a robust immunohistochemical marker has significantly improved diagnostics.[24] Georgiesh at al[25] aimed to investigate the clinicopathological and prognostic impact of different
fusion variants, and they found a correlation between a variant of NAB2-STAT6 fusion
and an increased risk of recurrence. Anyway, the prognostic value of fusion variants
should be further explored, in large patient cohorts. This gene fusion has not been
investigated in our patients. It would be good to investigate this gene fusion and
its prognostic value in future studies.
In our study, 20 patients had recurrence, most of them after more than 5 years from
surgery. None of them had metastasis. Histological features of malignancy and higher
preoperative maxSUV were significantly related to recurrence.
There was no difference in recurrence rate between SFTPs originating from visceral
or parietal pleura. Disease-free interval suggested that patients diagnosed and surgically
treated for SFTP, in particular patients with malignant form, need a long annual follow-up
period, in order to detect tumor recurrence early and to perform surgery rapidly.
The treatment of recurrence was surgical in all cases except two, because of their
poor general conditions. The surgical approach was mainly open through lateral thoracotomy,
since they were all reintervention. The excision was associated to a lung resection,
wedge in 15 cases and lobectomy in 3 cases.
Third, the prognosis for patients with SFTP is favorable with a 5- and 10-year OS rate
of 95 and 70%, respectively. Recurrence is frequent for this kind of tumor, with a
mean DFS of 143.3 ± 6.1 months. To date, the role for adjuvant therapy remains controversial.
Chemotherapy and radiotherapy are not largely used and data available in literature
are limited. Bylicki et al[26] highlighted the importance of the De Perrot classification as prognostic factor
and used it as a guideline for adjuvant therapy: they gave postoperative chemotherapy
based on doxorubicin to patients with stage III/IV cancer. Despite it, the majority
of these patients presented tumor recurrence. In the same study, chemotherapy for
recurrence had a modest response. Suter et al[27] described long-term survival using adjuvant radiotherapy in patients in whom an
R0 resection was not obtained. In our experience, all recurrences were treated only
with a surgical approach. Long-term survival after a second intervention for recurrence
was excellent, with 95 and 70% of patients alive after 5 and 10 years, respectively.
One of them died from causes related to the SFTP.
Fourth, according to our experience, the gold standard treatment of SFTPs is a radical surgical
excision. The surgical approach, VATS or thoracotomy, varies from case to case on
the basis of the localization of the tumor, its dimensions, and its relations with
the surrounding structures. Recurrence is frequent in these tumors. It can present
even more than 5 years after the surgical resection. In our experience, recurrence
is related to the histological features of malignancy and to maxSUV. Even benign tumors,
when presented with higher maxSUV, had a higher risk of recurrence. Since the gold
standard treatment of recurrence remains surgical excision, an early detection is
necessary to obtain a radical excision of the recurrence. We propose a long-term follow-up
with a CT scan every 6 months for the first year and every year for the next years,
with particular attention for patients with higher maxSUV at PET-CT scan and with
histological features of malignancy.
Conclusion
In our experience, histological features of malignancy and maxSUV are significantly
related to recurrence, which can occur even years after the first diagnosis. In patients
with preoperative diagnosis of malignancy and higher maxSUV, it should be indicated
to perform an enlarged resection with margins of more than 2 cm, in order to prevent
recurrence. Surgical excision with negative surgical margins results in good long-term
outcomes. After surgery, a long-term and strict follow-up should be done, in order
to detect recurrence early. R0 resection of a recurrent tumor, when feasible, should
be performed because it is associated with long-term survival.
Study Limitations
Our study presented several limitations that should be considered before drawing definitive
conclusions. The retrospective nature and the limited number of patients were the
main limitations. Also, we did not focus on NAB2-STAT6 fusion. In future studies,
it could be interesting to also focus on the molecular implications of NAB2-STAT6
fusion to find the prognostic value of fusion variants. Another limitation is that
non-R0 resected SFTs were excluded from analysis. It could also be interesting to
investigate the correlation between R1–R2 resection and recurrence.
Erratum: An erratum has been published for this article (DOI: 10.1055/s-0044-1791983).
