|Home | About | Journals | Submit | Contact Us | Français|
We sought to define the long-term outcome of surgically treated solitary fibrous tumors of the pleura.
We performed a retrospective review from December 1972 through December 2002.
There were 84 patients (39 men and 45 women) with a median age of 57 years (range, 34–83 years). Forty-six patients were symptomatic. Surgical resection included pulmonary wedge excision in 62 patients, lobectomy in 4 patients, segmentectomy in 2 patients, chest wall resection in 3 patients, isolated pleural resection in 7 patients, and chest wall resection with pulmonary wedge excision, lobectomy, or pneumonectomy in 3, 2, and 1 patients, respectively. Tumors were polypoid in 57 patients, sessile in 20 patients, and intrapulmonary in 7 patients. Histopathology was benign in 73 and malignant in 11 patients. Nine (82%) patient with malignant tumors and 37 (54%) patients with benign tumors were symptomatic (P = .11). The median tumor diameters for malignant and benign tumors were 12.0 and 4.5 cm, respectively (P = .001). Operative mortality and morbidity occurred in 3 (3.6%) and 7 (8.1%) patients, respectively. Median follow-up in survivors was 146 months (range, 23–387 months). Median survival for patients with benign and malignant tumors was 284 and 55 months, respectively, and 5-year survival was 88.9% and 45.5%, respectively (P = .0005). Eight (9.5%) patients had recurrent solitary fibrous tumors of the pleura. Recurrences were malignant in 6 and benign in 2 patients. Localized chest recurrences occurred in 3 patients, all of whom had reresection, with 2 patients again having recurrence.
Resection of benign solitary fibrous tumors of the pleura carries an excellent prognosis. Larger tumors are more likely to be malignant. Both benign and malignant tumors can recur. Although prolonged survival after resection of malignant tumors is possible, recurrence is common.
Solitary fibrous tumors of the pleura (SFTPs) are rare mesenchymal tumors most commonly arising in the pleura. They can, however, also present as an intrapulmonary mass.1 SFTPs are distinct from diffuse malignant mesothelioma (DMM), which is associated with asbestos exposure and typically carries a poor prognosis.2 The histopathologic characteristics of SFTPs were first described by Klemperer and Rabin in 1931.3 Over the years, SFTPs have been characterized by a number of terms, including fibrous mesothelioma, localized fibrous mesothelioma, localized mesothelioma, localized fibrous tumors, fibrous mesothelioma, solitary fibrous mesothelioma, pleural fibroma, submesothelial fibroma, and subserosal fibroma.4 The absence of clear nomenclature has only added to the confusion about these tumors, and uncertainty exists with regard to their clinical behavior. The purpose of this report is to review our experience with patients who underwent surgical resection for SFTPs.
Between December 1972 and December 2002, 84 patients underwent surgical treatment for SFTPs at the Mayo Clinic in Rochester, Minnesota, and these patients formed the basis for this review. The study was approved by the Mayo Foundation’s Institutional Review Board. Medical records were analyzed for age, sex, signs and symptoms, radiographic findings, tumor histopathology, location, surgical resection, clinical outcome, and survival. Operative mortality included patients who died within the first 30 days after surgical intervention or during the same hospital admission. Follow-up was obtained from retrospective review of the patients’ medical records and, where necessary, written surveys. Malignant SFTPs were diagnosed when more than 4 mitoses per 10 high-power fields were seen on light microscopic analysis.5 Additional microscopic supportive features of malignancy included areas of hemorrhage and necrosis, areas of high cellularity with crowding and overlapping nuclei, marked pleomorphism, and vascular invasion.5 The probability of survival and the cumulative probability of recurrence were estimated by using the Kaplan–Meier survivorship method.6 Recurrence and survival were estimated starting at the date of the first SFTP resection and are reported with 95% confidence intervals (CIs). Survival curves of the patients were compared with expected curves of persons of the same age and sex, as derived from vital statistics for the west–north–central region of the United States.7 Comparisons of survivorship curves were made with 1-sample log-rank tests.8 Comparisons of discrete variables between benign and malignant pathology were made with Fisher’s exact tests.9 Comparisons of continuous variables between benign and malignant pathology were made with Wilcoxon rank sum tests.10 All statistical tests were 2-sided.
There were 45 women and 39 men with a median age of 57 years (range, 34–83 years) enrolled in the study. Forty-five (54%) patients had a positive history of cigarette smoking. Median pack-years were 30 (range, 1–100 pack-years). Forty-six (55%) patients were symptomatic, with chest pain in 15 (18%), cough in 8 (9.5%), dyspnea in 3 (3.6%), and various other symptoms in 12 (14%). Eight (9.5%) patients had multiple symptoms. Extrathoracic symptoms, such as hypertrophic pulmonary osteoarthropathy alone in 3 (3.6%) patients, hypertrophic pulmonary osteoarthropathy and digital clubbing in 5 (6.0%) patients, and hypoglycemia in 1 (1.2%) patient, were present. Chest radiographs were obtained for the entire study group, and the SFTPs were identified in 77 (92%) patients. Before use of computed tomographic (CT) analysis, conventional chest tomographic scans were obtained in 15 (18%) patients, and the SFTPs were visualized in all. Sixty-two (74%) patients had CT scans of the chest, with the SFTPs visualized in 61 (98%). Magnetic resonance imaging was obtained in 4 (4.8%) patients, and results were abnormal in all. The SFTPs occurred on the right side in 50 (60%) patients and the left side in 34 (40%) patients. CT-guided percutaneous biopsies were performed in 22 (26%) patients and were diagnostic in 10 (45%) patients. Transbronchial biopsies were performed in 3 (3.6%) patients and were diagnostic in 1 (33%) patient.
All patients underwent surgical treatment. Postoperative chemotherapy and chemoradiotherapy were administered to 1 patient each. The operative approach was a posterolateral thoracotomy in 68 (81%) patients, video-assisted thoracic surgery (VATS) in 12 (14%) patients, a sternotomy in 3 (3.6%) patients, and a transabdominal approach in 1 (1.2%) patient. The transabdominal approach was used for a paraesophageal hiatal mass, which at the time of surgical intervention proved to be an SFTP that was able to be resected through this approach. The SFTPs were polypoid in 57 (68%) patients, sessile in 20 (24%) patients, and intra-pulmonary in 7 (8.3%) patients. At the time of the operation, the SFTPs appeared to arise from the visceral pleura in 58 (69%) patients, the parietal pleura in 15 (18%) patients, and both in 4 (4.8%) patients and were intrapulmonary without obvious visceral pleural involvement in 7 (8.3%) patients.
Surgical resection was pulmonary wedge excision in 62 (74%) patients, lobectomy in 4 (4.8%) patients, segmentectomy in 2 (2.4%) patients, isolated parietal pleural resection alone in 7 (8.3%) patients, chest wall resection alone in 3 (3.6%) patients, and chest wall resection combined with concomitant pulmonary wedge excision, lobectomy, or pneumonectomy in 3, 2, and 1 patients, respectively. SFTP resection was complete in all patients. SFTPs were benign in 73 (87%) patients and malignant in 11 (13%) patients. Nine (82%) patients with malignant SFTPs and 37 (54%) patients with benign SFTPs were symptomatic (P = .11). Morphologically benign SFTPs were polypoid, sessile, and intraparenchymal in 51 (70%), 15 (20%), and 7 (10%) patients, respectively. Similarly, malignant SFTPs were polypoid and sessile in 4 (36%) and 7 (64%) patients, respectively. Morphologic and pleural characteristics of the SFTPs are shown in Table 1. Neither the morphologic characteristics nor the site of pleural origin aided in the differentiation of benign from malignant SFTPs (P =.08). The median overall tumor diameter was 5 cm (range, 0.8–30 cm). The median diameter was 12 cm (range, 7–19 cm) for malignant SFTPs and 4.5 cm (range, 0.8–30 cm) for benign SFTPs (P =.001). In 7 (64%) of the 11 patients with malignant SFTPs, the tumor diameter was greater than 10 cm, whereas the tumor was greater than 10 cm in only 17 (23%) of the 73 patients with benign SFTPs (P = .01).
Mediastinal lymphadenectomy or lymph node sampling was performed at the individual discretion of the surgeon in 11 (13%) patients. Four of these patients had recurrent SFTPs, and 6 had malignant tumors. Mediastinal lymph node removal was undertaken on larger SFTPs with a median diameter of 7.5 cm (range, 1.5–19 cm). On histopathologic examination, no tumor was found in any of the lymph nodes. Complications occurred in 9 (11.8%) patients and included atrial fibrillation, bleeding, and reoperation for bleeding in 2 (2.4%) patients each and pleural effusion, pneumonia, and postoperative ileus in 1 (1.2%) patient each.
Three patients died (operative mortality, 3.6%). One patient with no prior history of coronary artery disease died suddenly on postoperative day 3 of a myocardial infarction. The second patient had known renal insufficiency and severe atherosclerosis and died on postoperative day 9 of mesenteric ischemia. The final patient had been discharged without complication from the hospital on postoperative day 7 and died of unknown cause on postoperative day 12 at home.
Follow-up was complete in all survivors and ranged from 2 to 32 years (median, 12 years). At last follow-up, 56 patients were alive, and 27 had died. The overall 5-year survival was 83% (95%CI, 75%–92%), with a median survival of 284 months and an expected survival of 91%(P = .19, Figure 1). Median survival for malignant SFTPs was 55 versus 284 months for benign SFTPs. Five-year survival for benign and malignant SFTPs was 89% (95% CI, 82%–96%) and 45.5% (95% CI, 24%–87%), respectively (P = .0005, Figure 2).
Overall, 5-year freedom from recurrence was 95% (95% CI, 90%–100%). Of 81 operative survivors, 8 (9.9%) patients experienced recurrent SFTPs. For patients experiencing recurrence, the surgical approach was a thoracotomy in 7 patients and a median sternotomy in 1 patient. The site of pleural origin for recurrent SFTPs was visceral in 4 patients, parietal in 2 patients, and both in 2 patients. Five-year freedom from recurrence for patients with benign and malignant SFTPs was 100% and 58% (95% CI, 34%–100%), respectively (P < .0001, Figure 3). Ultimately, 6 (75%) patients with malignant SFTPs and 2 (25%) patients with benign SFTPs experienced recurrence. Median time to recurrence for malignant and benign SFTPs was 34 and 83 months, respectively. In the 6 patients with initially malignant SFTPs that recurred, the recurrences were malignant in 5 patients and benign in 1 patient. In the 2 patients with benign SFTPs that recurred, the recurrence was benign and malignant in 1 patient each. Recurrences were localized in 3 (37.5%) patients and widely metastatic in 5 (62.5%) patients. All 3 patients with localized recurrences underwent reresection, and 2 STFPs again recurred. The SFTP recurrence was diffusely metastatic within the abdominal cavity and localized to the ipsilateral hemithorax in 1 patient each. The latter patient eventually underwent a total of 4 resections over a 24-year period, after which he remained without evidence of disease. In the 5 patients with widely metastatic recurrent SFTPs, the observed sites of distant metastases included bone, contralateral hemithorax, liver, peritoneum, and retroperitoneum.
SFTPs are uncommon neoplasms, with just over 900 cases reported in the literature.4,11 Although SFTPs were historically considered benign tumors, Okike and colleagues,12 from our institution, in 1978 emphasized their malignant potential. Nevertheless, because of a lack of understanding of the cellular origins of SFTPs and their seemingly indolent clinical course, many physicians still consider SFTPs to be a benign variant of mesothelioma.13
Localized mesothelioma of the pleura was first described by Wagner in 1870.12 In 1931, Klemperer and Rabin3 distinguished localized mesothelioma from diffuse mesothelioma and proposed submesothelial cells as the cellular origin for the localized variety. Controversy developed in 1941, when Stout and Murray14 claimed mesothelial cells to be the cellular origin of localized mesothelioma. More recently, immunohistochemical techniques, flow cytometric studies, and electron microscopic analysis have demonstrated that SFTPs indeed arise from the mesenchymal cells beneath the mesothelial lining of the pleura, as initially postulated by Klemperer and Rabin.2–4,11,15,16 This mesothelial versus mesenchymal difference in cell origin is likely key to the wide dissemination of DMM over serosal surfaces, eventually leading to organ encasement, in contrast to the more localized spread of SFTPs. Based on this differentiation in cellular origin, the term localized mesothelioma has been abandoned in favor of SFTPs.17 SFTPs have no known association with asbestos exposure. DMM, unlike SFTPs, has a poor clinical course, with death often occurring within 2 years of diagnosis. Immunohistochemistry is a useful adjunct in the differentiation of SFTPs from DMM. SFTPs, in distinction to DMM, stain positive for vimentin and CD34 and negative for keratin. Table 2 outlines the distinct immunohistochemical profile of SFTPs and DMM.
Our series, like others, show SFTPs to most commonly arise in the sixth decade of life. Although some authors have stated that SFTPs are seen equally in men and women,4,15 our large series and others11,12,18 show a predilection for women. Approximately half of patients are symptomatic on presentation, and symptoms are more common in patients with malignant SFTPs. Common symptoms include chest pain and cough. Extrathoracic symptoms, such as hypertrophic pulmonary osteoarthropathy, clubbing, and hypoglycemia, have been reported in up to 66% of patients12,19,20 but occurred in only 9 (11%) of our patients. This frequency of symptoms is less than both the 66% and 19% frequencies reported in our earlier series12,19 but still greater than the frequencies reported in most recent series from other institutions.11,16,18 The extrathoracic symptoms resolved in all of our patients after surgical resection. Chest CT scanning is the radiographic test of choice, often showing a well-circumscribed homogenous mass of soft tissue attenuation; however, this finding lacks specificity, and heterogeneity is possible (Figures 4 and and5).5). The tumors commonly show displacement of surrounding structures, although some might appear as intraparenchymal masses. CT scanning cannot differentiate benign from malignant SFTPs. The diagnostic accuracy of CT-guided biopsy was 45% in our series and thus, like other series, is considered unacceptably low.13,16,18,21 Because of these variable results, we do not routinely recommend CT-guided biopsy. CT-guided biopsy would be considered in patients at too high a risk for surgical resection or patients with radiographically unresectable masses to try and obtain a histopathologic diagnosis.
Complete surgical resection is the preferred treatment for SFTPs. VATS was used in only 14%of our patients, but this clearly reflects the 30-year time span of this series, with VATS being available only in the latter 15 years. The median SFTP size for VATS resections was 2.4 cm (range, 0.8–15 cm), and there were no recurrences. The median SFTP size for open resections was 7 cm (range, 1–30 cm). Pulmonary wedge excision was the most common surgical procedure, occurring in nearly three quarters of our patients, and formal anatomic lung parenchymal resections alone were performed in only 6 (7.1%) patients. Chest wall resection alone was required in 3 (3.6%) patients and was associated with an additional lung parenchymal resection in 6 more patients. The operative approach and extent of surgical excision should be dictated by the SFTP’s size and location. SFTPs in our series ranged in size from 0.8 to 30 cm, with a median diameter of 5 cm. Size was the best clinical predictor of malignancy, with malignant tumors having a median diameter of 12 cm in comparison with 4.5 cm in benign tumors (P = .001).
Long-term survival after resection of both benign and malignant SFTPs is possible, with the overall median survival in our series being 24 years. However, long-term survival for completely resected benign SFTPs is dramatically better than that of completely resected malignant SFTPs (Figure 2). Long-term follow-up after complete resection is necessary because SFTPs can recur, and this was seen in 8 (9.9%) of our patients, including 2 patients with benign histopathology. Several authors have not reported recurrence after complete resection of benign SFTPs.11,13,16,18,21 In contrast, reported recurrence rates for completely resected malignant SFTPs range from 14% to 86%.11,13,16,18 In our series there were 10 surviving operative patients with malignant SFTPs, and 6 (75%) of these experienced recurrence. In only 3 patients (2 with benign SFTPs and 1 with a malignant SFTP) was the recurrence localized to the same hemithorax. After reoperation, recurrence again occurred in 2 patients.
In conclusion, SFTPs are uncommon tumors that are best treated with complete surgical resection. In this series increased tumor size is suggestive of malignancy. Long-term survival after resection of benign SFTPs is excellent but is also possible after resection for malignant SFTPs. However, malignant SFTPs are prone to recurrence, which in turn is associated with a poor prognosis.
Read at the Eighty-seventh Annual Meeting of the American Association for Thoracic Surgery, Washington, DC, May 5–9, 2007.
Disclosures: Dr Stephen D. Cassivi reports consulting fees from Spiration Inc.