From January 1983 to December 2013, 26 patients with SFTP were identified. There were 18 (69.2%) female and 8 (30.8%) males, with mean age of 63.6 years (range 41.6 to 83.1). Mean follow-up after initial treatment was 125.8 months (range 0.25 - 581). Baseline population demographics are summarized in Table 1. The majority of patients were asymptomatic at presentation (n = 15, 57.7%), with the pleural mass identified incidentally by plain chest X-ray conducted for an unrelated reason. Dyspnea on exertion was the most common presentation in those with symptoms. A minority of patients exhibited paraneoplastic symptoms. Further data on clinical presentation is summarized in Table 2.

Prior to determining initial management, a trans-thoracic needle of the pleural mass was conducted in 22 (84.6%), the majority of cases (Table 3). The biopsies were diagnostic of SFTP in 16 (72.7%), and 3 (13.6%) of these diagnosed malignant SFTP. Of the 3 diagnoses of malignant pathology on pre-management biopsy, 1 was consistent on final surgical pathology with malignant SFTP and the diagnosis was revised in the other two as benign SFTP. Of the 13 other cases where benign or malignant SFTP could not be determined conclusively, 5 additional cases were deemed of malignant potential on final surgical pathology.

Primary management was surveillance in 3 and complete surgical resection in 23 (Table 4). Preoperative evaluation in the majority included CT thorax and pulmonary function tests. PET scan, bone scan and further cardiopulmonary assessment was conducted selectively on a case by case basis. Adjuvant chemotherapy or radiotherapy was not utilized in any, however 6 (26.2%) cases were discussed at multidisciplinary thoracic cancer conference for consideration of adjuvant therapy. A total of 4 cases experienced pathologically confirmed recurrence following complete surgical resection: 3 isolated local, & 1 distant recurrence. The initial surgical pathology was reported as benign SFTP in 3 (75%) and malignant SFTP in 1 (25%). Pertinent descriptive details regarding recurrent disease are displayed in Table 5.

MCC: Multidisciplinary Thoracic Cancer Conference. SD: Standard deviation. SFTP: Solitary fibrous tumor of the pleura.

*Does not include biopsies done to diagnose recurrent SFTP disease. SD: Standard deviation. SFTP: Solitary fibrous tumor of the pleura.

SFTP: Solitary fibrous tumor of the pleura. MCC: Multidisciplinary Thoracic Cancer Conference.

*Times to local recurrence or metastasis include patients with known pathologically proven recurrence/metastasis only. **Sites of metastasis includes all sites in a single patient with progressive multiple metastatic disease. MCC: Multidisciplinary Thoracic Cancer Conference. SD: Standard deviation. SFTP: Solitary fibrous tumor of the pleura.

The relationship between recurrent disease and demographic, clinical, pathological and immunohistochemical variables is summarized in Table 6. In keeping with previous series, none of the variables analyzed were predictive of SFTP recurrence following complete surgical resection (including tumor size, malignant initial histology and reported immunohistochemical stains).

Inferential analysis of clinical, pathologic and immunohistochemical variables for their relationship with reported malignant pathologic potential did reveal several significant associations (Table 7). Mitoses >4/10 high power fields (p = 0.003), presence of tumor necrosis (p = 0.001), High cellularity (p < 0.001), presence of nuclear pleomorphosim (p < 0.001), nuclear crowding/overlapping (p = 0.006), and an infiltrative pattern of cells in relation to surrounding structures (p = 0.042), were all associated with SFTP of high malignant pathologic potential. The association with nuclear crowding and infiltrative cell pattern has not been reported previously. There was also a trend observed for larger mean tumor diameter to be associated with malignant pathology (p = 0.070), in particular, tumors >5 cm in diameter (p = 0.074). Immunohistochemical marker expression did not correlate with malignant pathology.

*t-test for comparison of means or Fisher exact test for comparison of binary categorical variables. **Tumor size by CT Thorax or Surgical pathology. CT Thorax used for tumor size only if primary management was not surgical.

Table does not include biopsies or surgery for recurrent disease. *t-test for comparison of means, Fisher exact test for comparison of binary categorical variables.

The majority of patients in our series received an image-guided trans-thoracic needle biopsy 22(84.6%) prior to initial management, and of these the majority were diagnostic of SFTP 16/22 (72.7). Thus the role of biopsy in this setting remains diagnostic, namely to distinguish SFTP from other pleural neoplasms on the differential diagnosis requiring a different management strategy. We found that trans-thoracic fine needle aspiration biopsy provided tissue for histological analysis of cellular & nuclear features of malignancy (mitotic count, nuclear pleomorphism, nuclear crowding, etcetera), but may not provide sufficient tissue for definitive diagnosis based on immunohistochemistry (CD34, bcl-2, etcetera) (Table 3). As such, we recommend that a core needle biopsy be completed concurrently with FNA in order to provide sufficient tissue for IHC analysis and SFTP diagnosis. Given the unpredictable potentially aggressive behavior of these tumors (discussed in next section below), a diagnosis of SFTP on biopsy without observed malignant pathological features is not sufficient to call the tumor “benign”. A more apt description in this scenario is “SFTP with uncertain malignant potential”. For definitive diagnosis of malignant potential, while simultaneously providing optimal tumor management, we recommend surgical resection.

The findings regarding predictors of recurrent disease in our study were in keeping with previous of series [16] [18] [23] [24] . Aggressive behavior in these tumors is unpredictable. Despite analysis of a large range of clinical, histopathological and immunohistochemical variables, there were no significant predictors of recurrent disease identified (Table 5). We also showed that 3 or our 4 cases with recurrent disease initially displayed “benign” surgical pathology following complete resection. We did identify several predictors of “malignant pathology” based on accepted histopathological features (Table 7). However, statistical comparison of the relationship between benign or malignant final surgical pathology did not reveal a significant association with tumor recurrence. This is particularly concerning regarding the concept of “benign” SFTP nomenclature, and suggests that all tumors are best considered potentially malignant. It may be most prudent to describe these tumors in pathological terms of “lower malignant potential” and “higher malignant potential” instead of “benign” and “malignant.”

Long-term clinical follow-up with imaging, regardless of initial benign or malignant final pathological diagnosis, is required for these tumors given their documented unpredictable aggressive behavior. The best imaging modality (CT thorax, Chest X-ray), optimal surveillance intervals, and which clinician is best suited to conduct such follow-up, remain unclear.

Limitations in any study of SFTP arise from its rarity. In our study, despite a 30 year review a small study sample remained. This necessitated consideration of a potentially underpowered small sample size when conducting statistical analysis and interpreting the results. Additional limitations arose from the retrospective nature of the series and associated reporting bias. In particular, the promising role of bFGF and Ki-67 immunohistochemical labeling indices role in prognostication of recurrence could not be evaluated due to limited reporting.

Standards of practice and nomenclature for this tumor also evolved over the 30 year study period. Accepted tumor nomenclature changed following the rise in knowledge of the mesenchymal origin for these neoplasms from benign mesothelioma, to SFTP, to benign or malignant SFTP [1] [2] . Practice and diagnostic capability also differed over the study period with development of new technology and expertise in IHC marker interpretation. Finaly, despite evolving diagnostic pathological expertise, accepted criteria to determine benign or malignant SFTP pathology remains unstandardized.