The American Cancer Society predicts 10,520 new cases and 3920 deaths from soft tissue sarcoma (STS) for 2010. STS disseminates primarily via the hematogenous route, although lymphatic spread does occur with certain subtypes. The lung is the most common metastatic site in most large series, accounting for up to 80% of metastases. The median overall survival for pulmonary metastatic disease with current multidisciplinary treatment is approximately 12 to 14 months. Pulmonary metastasectomy (PM) represents the only potentially curative treatment for patients with STS and lung metastases. This article discusses the management of STS using PM.
The American Cancer Society predicts 10,520 new cases and 3920 deaths from soft tissue sarcoma (STS) for 2010. STS disseminates primarily via the hematogenous route, although lymphatic spread does occur with certain subtypes. The lung is the most common metastatic site in most large series, accounting for up to 80% of metastases. The rate of metastatic disease primarily depends on tumor grade and ranges from 10% for low grade to 30% for intermediate and greater than 50% for high grade. Approximately 70% of pulmonary metastases occur within the first 2 years after primary tumor resection.
The median overall survival (OS) for pulmonary metastatic disease with current multidisciplinary treatment is approximately 12 to 14 months. Pulmonary metastasectomy (PM) represents the only potentially curative treatment for patients with STS and lung metastases. The first PM was described by Weinlechner in 1882 when 2 lesions were removed incidentally during resection of a chest wall sarcoma. Alexander and Haight were the first to show a survival benefit and describe criteria for resection, which include control of the primary tumor, absence of extrathoracic disease, and sufficient pulmonary reserve. These criteria remain valid to this day.
PM is the accepted standard therapy for a small subset of patients with STS with resectable disease and favorable tumor biology. This standard is based on an abundance of retrospective studies showing improved survival with complete resection versus historical controls without resection. Despite aggressive surgical management, recurrence rates are still greater than 50%.
Imaging evaluation
Ideally, chest imaging identifies all metastatic lesions and is able to differentiate benign from malignant processes. However, inflammatory conditions can often mimic metastases. Pulmonary metastases are usually described as multiple round nodules of varying size with indistinct borders, spiculation, pleural thickening, and inhomogeneity of the surrounding lung parenchyma. Certain sarcomas commonly present with calcified metastases (osteogenic/chondrogenic and synovial sarcoma) and can be mistaken for benign processes such as granulomatous disease.
Chest radiography remains an important modality for the staging and surveillance of patients with STS. The low cost, universal availability, and excellent safety record are the reasons for its continued utility. Despite being very specific for the identification of pulmonary metastases, sensitivity is limited.
Computed Tomographic Scan
The improved sensitivity of computed tomographic (CT) scans over plain chest radiography has framed a more recent debate regarding the best screening modality for pulmonary metastases. The advent of standard CT imaging followed by helical CT with thin sectioning has increased the detection rate (sensitivity) for small pulmonary nodules. Sliding thin-slab maximum intensity projection (MIP) techniques create 2-dimensional representations of the initial thin sections to differentiate pulmonary vessels from nodules (by showing the whole course of the vessel within the slice), thus decreasing false-positives. The addition of computer-aided diagnosis (CAD) also increases sensitivity of pulmonary nodule detection as compared with standard CT scanning.
Although the addition of MIP or CAD to multidetector CT has shown an increased sensitivity for nodule detection, few studies have demonstrated a positive impact of these aids in the setting of pathologically proved malignancy. Of the few studies available, all but one deal with lung cancer. The lone study of both MIP and CAD in patients undergoing PM showed increased sensitivity per patient and per nodule. This increased sensitivity was valid only for nodules that are 5 mm or smaller. Colorectal, renal, and hepatocellular cancers comprised two-thirds of the patients, so this study may not be applicable to patients with STS.
CT scanning demonstrates pulmonary nodules in 33% of newly diagnosed patients with STS, and approximately 80% of these turn out to be metastases. Indeterminate lesions 10 mm or smaller make up 21% of nodules, of which 28% are metastatic. Patients having only lesions smaller than 5 mm that remain stable over 6 months enjoy similar survival as those with normal CT scan results. Magnetic resonance imaging has a sensitivity similar to that of conventional CT scan, so offers no specific advantage.
Positron Emission Tomography
Molecular imaging using positron emission tomography (PET) with fludeoxyglucose F 18 ( 18 F FDG-PET) is a useful modality for the differentiation of benign from malignant soft tissue masses. Moreover, it is also a useful tool to evaluate response to therapy. 18 F FDG-PET uptake can also predict prognosis and grade of the primary tumor. However, the sensitivity of 18 F FDG-PET for STS pulmonary metastases is relatively low (50%–86.5%) than that of CT (95.1%–100%). Specificity is similar for both modalities. Because of the lack of spatial resolution of 18 F FDG-PET, its fusion with CT into an integrated PET/CT improves sensitivity and specificity beyond what either achieves alone in malignancies such as lung cancer. Iagaru and colleagues examined 106 patients with osseous sarcoma and STS who underwent dedicated chest CT. Seventy-six of these patients underwent a dedicated PET, and the remaining 30 patients had hardware-fused PET/CT imaging. There was no difference in sensitivity or specificity between the 2 groups, concluding that PET/CT does not add power to diagnostic value over PET alone. Although PET/CT imaging may have improved since this study, PET/CT for STS is challenged by the lack of intravenous contrast and single breath-hold technique for the CT portion of the technology. Furthermore, lower-grade sarcomas do not have reliably high glucose metabolism (standardized uptake value) like malignancies for which PET/CT has proved superior.
Factors affecting patient selection for PM
PM represents the only potentially curative treatment for patients with STS and lung metastases. Despite aggressive surgical management, recurrence rates remain greater than 50%. Therefore, careful consideration is necessary when determining whether or not a patient will benefit from surgical intervention. Appropriate patient selection requires that the following criteria be confirmed, within reason. First, the primary tumor is controlled or controllable. Second, extrathoracic disease is absent. Third, there is sufficient pulmonary reserve to tolerate complete resection of all disease. Last, complete resection is achievable.
Synchronous Pulmonary Metastases
The proportion of patients with STS who initially present with synchronous pulmonary metastases is approximately 10%. There are only 2 studies specifically addressing patients with STS and synchronous disease. Kane and colleagues reviewed 48 patients with synchronous metastatic disease, of which 30 had pulmonary metastases. Thirteen of these patients underwent PM, and there was no significant difference in median survival between patients who underwent resection and who did not. Ferguson and colleagues studied 112 patients who presented with synchronous metastases. Eighty-eight (79%) patients had pulmonary metastases, 18 of whom underwent PM. Median OS was 9 months for all patients presenting with synchronous pulmonary metastases. There were 3 long-term survivors from the 18 patients who underwent resection, but this did not show significance on multivariate analysis. A similar study by Liebl and colleagues did not specifically look at synchronous presentation but included it in a univariate analysis. Patients with synchronous disease had a median survival of 21 months versus 40 months with metachronous metastases The poor outcome for synchronous metastases could be predicted, considering the known negative impact of short disease-free interval (DFI) in STS PM series. Given the poor results of PM in these cases, patients with synchronous metastatic disease should instead be considered for clinical trials or initial chemotherapy to assess response and tumor biology.
Impact of Primary Tumor Local Recurrence on PM
Limited data exist regarding the impact of prior local recurrence (LR) on outcome after subsequent PM. In Pisters and colleagues’ large series of patients with STS, the overall rate of distant metastatic disease was 22%. About 30% of these patients developed metastases synchronously with or after an LR. However, LR did not significantly affect postmetastasis survival. In Billingsley and colleagues’ study, 18% of patients with metastasis had an LR as a preceding event and another 6% had metastases synchronous with an LR. In contrast to Pisters and colleagues’ cohort, patients with an LR were twice as likely to die from their metastases. Chen and colleagues examined their experience with PM in 23 patients with STS who underwent complete resection. Multivariate analysis showed that 10 patients who had a prior LR were 20 times more likely to die than patients without a prior recurrence. Ueda and colleagues also found a better 5-year OS in patients without an associated LR (30.0% for LR vs 12.5% for no recurrence). This was significant on univariate analysis but was not subjected to multivariate analysis. Although these are only small series on which to base decisions, the presence of an LR may diminish the anticipated therapeutic benefit of PM.
Extrathoracic Metastatic Disease
Studies on patients with a combination of pulmonary and extrathoracic metastases are also limited. In a large series of extremity STS with distant metastases, nonlung metastases had a postresection survival similar to pulmonary metastases. In Blackmon and colleagues’ study, 36.2% of patients undergoing PM had prior synchronous or metachronous extrathoracic metastases. The most common sites of extrathoracic metastases were spine, bone, soft tissue, liver, abdomen, brain, and pelvis. Three groups were compared: pulmonary metastases only, pulmonary metastases plus synchronous or prior extrathoracic metastases, and pulmonary metastases plus subsequent metachronous extrathoracic metastases. There was no difference in survival between the group with resected pulmonary tumor only and that with resected pulmonary plus synchronously or previously resected extrathoracic tumor. Survival was very poor in patients undergoing PM in whom the extrathoracic metastases could not be completely resected. This finding supports resection of pulmonary metastases and extrathoracic metastases only when the tumor can be completely resected. In this situation, one could expect survival similar to isolated pulmonary metastases.
Adequate Pulmonary Reserve
In general, PM is associated with low morbidity and mortality rates. Patients with STS lung metastases do not typically have chronic obstructive pulmonary disease or other cardiopulmonary comorbidities common in patients with lung cancer. In addition, the PM surgical procedures are generally conservative wedge resections, and most lesions are peripherally located. Wedge resections or segmentectomies are performed in 73% to 86% of cases. Perioperative mortality is typically 0.0% to 3.7%. Perioperative morbidity ranges from 6% to 14%.
Completeness of Surgical Resection
Consistently, the factor most significantly associated with improved survival after PM is complete resection of the metastatic disease. In Smith and colleagues’ study, the median OS for R0 (complete) resection was 22 months, 11.5 months for R1 (microscopically positive margin) resection, and 9.5 months for R2 (grossly positive margin) resection. In Billingsley and colleagues’ study, the median survival after complete resection was 33 months versus 16 months for an incomplete resection.
Factors affecting patient selection for PM
PM represents the only potentially curative treatment for patients with STS and lung metastases. Despite aggressive surgical management, recurrence rates remain greater than 50%. Therefore, careful consideration is necessary when determining whether or not a patient will benefit from surgical intervention. Appropriate patient selection requires that the following criteria be confirmed, within reason. First, the primary tumor is controlled or controllable. Second, extrathoracic disease is absent. Third, there is sufficient pulmonary reserve to tolerate complete resection of all disease. Last, complete resection is achievable.
Synchronous Pulmonary Metastases
The proportion of patients with STS who initially present with synchronous pulmonary metastases is approximately 10%. There are only 2 studies specifically addressing patients with STS and synchronous disease. Kane and colleagues reviewed 48 patients with synchronous metastatic disease, of which 30 had pulmonary metastases. Thirteen of these patients underwent PM, and there was no significant difference in median survival between patients who underwent resection and who did not. Ferguson and colleagues studied 112 patients who presented with synchronous metastases. Eighty-eight (79%) patients had pulmonary metastases, 18 of whom underwent PM. Median OS was 9 months for all patients presenting with synchronous pulmonary metastases. There were 3 long-term survivors from the 18 patients who underwent resection, but this did not show significance on multivariate analysis. A similar study by Liebl and colleagues did not specifically look at synchronous presentation but included it in a univariate analysis. Patients with synchronous disease had a median survival of 21 months versus 40 months with metachronous metastases The poor outcome for synchronous metastases could be predicted, considering the known negative impact of short disease-free interval (DFI) in STS PM series. Given the poor results of PM in these cases, patients with synchronous metastatic disease should instead be considered for clinical trials or initial chemotherapy to assess response and tumor biology.
Impact of Primary Tumor Local Recurrence on PM
Limited data exist regarding the impact of prior local recurrence (LR) on outcome after subsequent PM. In Pisters and colleagues’ large series of patients with STS, the overall rate of distant metastatic disease was 22%. About 30% of these patients developed metastases synchronously with or after an LR. However, LR did not significantly affect postmetastasis survival. In Billingsley and colleagues’ study, 18% of patients with metastasis had an LR as a preceding event and another 6% had metastases synchronous with an LR. In contrast to Pisters and colleagues’ cohort, patients with an LR were twice as likely to die from their metastases. Chen and colleagues examined their experience with PM in 23 patients with STS who underwent complete resection. Multivariate analysis showed that 10 patients who had a prior LR were 20 times more likely to die than patients without a prior recurrence. Ueda and colleagues also found a better 5-year OS in patients without an associated LR (30.0% for LR vs 12.5% for no recurrence). This was significant on univariate analysis but was not subjected to multivariate analysis. Although these are only small series on which to base decisions, the presence of an LR may diminish the anticipated therapeutic benefit of PM.
Extrathoracic Metastatic Disease
Studies on patients with a combination of pulmonary and extrathoracic metastases are also limited. In a large series of extremity STS with distant metastases, nonlung metastases had a postresection survival similar to pulmonary metastases. In Blackmon and colleagues’ study, 36.2% of patients undergoing PM had prior synchronous or metachronous extrathoracic metastases. The most common sites of extrathoracic metastases were spine, bone, soft tissue, liver, abdomen, brain, and pelvis. Three groups were compared: pulmonary metastases only, pulmonary metastases plus synchronous or prior extrathoracic metastases, and pulmonary metastases plus subsequent metachronous extrathoracic metastases. There was no difference in survival between the group with resected pulmonary tumor only and that with resected pulmonary plus synchronously or previously resected extrathoracic tumor. Survival was very poor in patients undergoing PM in whom the extrathoracic metastases could not be completely resected. This finding supports resection of pulmonary metastases and extrathoracic metastases only when the tumor can be completely resected. In this situation, one could expect survival similar to isolated pulmonary metastases.
Adequate Pulmonary Reserve
In general, PM is associated with low morbidity and mortality rates. Patients with STS lung metastases do not typically have chronic obstructive pulmonary disease or other cardiopulmonary comorbidities common in patients with lung cancer. In addition, the PM surgical procedures are generally conservative wedge resections, and most lesions are peripherally located. Wedge resections or segmentectomies are performed in 73% to 86% of cases. Perioperative mortality is typically 0.0% to 3.7%. Perioperative morbidity ranges from 6% to 14%.
Completeness of Surgical Resection
Consistently, the factor most significantly associated with improved survival after PM is complete resection of the metastatic disease. In Smith and colleagues’ study, the median OS for R0 (complete) resection was 22 months, 11.5 months for R1 (microscopically positive margin) resection, and 9.5 months for R2 (grossly positive margin) resection. In Billingsley and colleagues’ study, the median survival after complete resection was 33 months versus 16 months for an incomplete resection.
Surgical approach to PM
Planning a PM for STS requires careful consideration of each case individually. Rather than relying on 1 approach, it is probably most advantageous to have all techniques available to fit to each unique situation. The ideal approach depends on the surgeon’s experience and the anatomic locations of the metastases. In addition, minimally invasive surgical options are being used with increasing frequency, representing a potentially curative and repeatable procedure with less pain, decreased hospital stay, and faster recovery.
Median Sternotomy
The main advantage of median sternotomy is simultaneous access to both pleural cavities. Another advantage over standard posterolateral thoracotomy is less pain and respiratory compromise. However, the heart limits exposure of the hila and left lower lobe. The more posterior segments are also less accessible. Although infrequent, sternal wound complications cause high morbidity.
Transsternal Thoracotomy
Bilateral anterolateral thoracotomies with transsternal extension (clamshell incision) offer optimal exposure, especially the hila and left lower lobe, as compared with median sternotomy. Although postoperative pain is greater than with median sternotomy, because of the sternal release, less force is applied to fewer rib interspaces and may lead to less pain than bilateral traditional thoracotomies. This approach does sacrifice the bilateral internal thoracic arteries and precludes their use for flap reconstruction or revascularization procedures.
Posterolateral Thoracotomy
A traditional posterolateral thoracotomy through the fifth or sixth interspace provides excellent exposure for unilateral and hilar lesions. There was initially some controversy surrounding performing unilateral thoracotomy for patients with unilateral disease by imaging. The National Cancer Institute reported a higher yield of occult pulmonary nodules for median sternotomy than unilateral thoracotomy. Despite finding more disease, survival was not improved. Younes and colleagues compared patients who underwent bilateral thoracotomies for bilateral disease with patients initially noted to have unilateral disease undergoing unilateral thoracotomy in whom the condition recurred in the contralateral lung. The investigators found that delaying contralateral thoracotomy until disease became radiologically apparent did not affect OS.
Video-Assisted Thoracoscopic Surgery
Experience with minimally invasive thoracic surgical approaches continues to grow as the techniques and supporting technology mature. Video-assisted thoracoscopic surgery (VATS) offers patients with STS and lung metastases a potentially curative and repeatable treatment option with low perioperative morbidity. The rationale for VATS comes from several factors identified in PM series. The initial concerns regarding VATS centered on the lack of manual palpation of the lungs like what is performed during open thoracotomy. Additional nonimaged but palpable nodules were identified in 42% of patients undergoing PM. In a recent series using multidetector CT scan with thin cuts and fused PET/CT, 37% of patients had missed nodules found with manual palpation; 18% had a missed malignant nodule. This view that VATS was an inferior approach because of missed metastases was further supported by results of a prospective trial of VATS followed by confirmatory thoracotomy showing that additional malignant lesions were identified in 56% of patients at confirmatory thoracotomy.
A contrary view of VATS arose from the knowledge that a significant neoplastic burden remains in micronodules (1–2 mm) that are beyond the detection of both CT and manual palpation. Consequently, the intrathoracic recurrence rate in open thoracotomy series has been as high as 50% to 69%. Even in a median sternotomy series in which both thoracic cavities were palpated, the intrathoracic recurrence rate was nearly 40%. Roth and colleagues also showed that occult contralateral disease existed in 40% of patients with STS undergoing PM via sternotomy as compared with unilateral thoracotomy, but the choice of unilateral exploration did not affect survival.
In series comparing VATS with thoracotomy, both modalities showed similar survival. Carballo and colleagues examined 171 patients with various malignancies (47% sarcoma). Of those, 135 underwent thoracotomy and 36 underwent VATS. The 5-year OS was 58.8% in the thoracotomy group and 69.6% in the VATS group ( Fig. 1 ). The subgroup with the highest survival included patients with sarcoma undergoing VATS. Gossot and colleagues examined STS PM with characteristics favorable for VATS: 2 or less lesions per lung field, maximum tumor dimension of 3 cm, wedge resection technically possible, and no extraparenchymal disease. A total of 31 patients underwent a VATS procedure and were compared with 29 patients with similar tumor characteristics who underwent standard thoracotomy. There was no difference in OS, disease-free survival (DFS), or ipsilateral recurrence rates. Mutsaerts and colleagues compared patients with a solitary nodule identified on preoperative CT prospectively undergoing VATS with similar historical controls who had undergone confirmatory thoracotomy. There was no difference in DFS or OS between the 2 groups at 2 years.
Potential benefits of VATS over standard thoracotomy include reduced postoperative morbidity, pain, decreased hospital stay, and earlier return to work. Another important potential benefit of VATS for STS PM is based on the role of repeat metastasectomy in some patients. In Weiser and colleagues’ study, 35% of their patients with STS undergoing an initial PM underwent a second thoracic exploration for recurrence and 37% of that subgroup underwent a third thoracic exploration. Kondo and colleagues specifically observed the benefit of VATS in repeat PM. Patients were placed into 4 groups: group A (previous VATS/present VATS), group B (previous VATS/present thoracotomy), group C (previous thoracotomy/present VATS), and group D (previous thoracotomy/present thoracotomy). Operative times, bleeding, and chest tube drainage in group D were significantly longer than those for the other 3 groups.
Size, depth, and consistency of the metastasis influence the ability to identify the nodule thoracoscopically. A variety of methods have been used to aid in the intraoperative identification of nodules, including intraoperative ultrasonography; percutaneous injection of dye, contrast material, or radionuclides; percutaneous hook wire; wire coil placement; and transbronchial localization. Preoperative localization is most helpful in the following situations: maximum diameter of nodule of 5 mm or less, maximum diameter to minimum distance between visceral pleura and inferior border of the nodule of 0.5 or less, low-density or ground-glass appearance of the lesion, and location in anatomically difficult areas such as basilar segments of the lower lobes. The most common method involves the use of the hook wire or suture system. This technique is highly successful with a low complication rate and has the added benefit over other techniques of offering a means of retraction intraoperatively.
Emerging technologies such as navigational bronchoscopy have been used to localize lung nodules for stereotactic radiosurgery, so it is conceivable that this technology could be used to aid in surgical excision also in the future.
Extent of resection
The goal of PM is a negative margin with preservation of lung parenchyma. Given the peripheral nature of STS metastases, this can often be accomplished with wedge resection. However, multiple lesions within a lobe or centrally located lesions may require more extensive resections.
Pneumonectomy
The International Registry of Lung Metastases reported on 5206 patients undergoing PM. Pneumonectomy was performed in 4% (171 patients). With pneumonectomy, there was a 4% preoperative mortality and a 5-year OS of 20% for patients who underwent complete resection (as compared with no survivors beyond 25 months in the group that underwent incomplete resection). The pneumonectomy group was dominated by patients with STS who had a 5-year OS of 30% if they underwent complete resection. Long-term survival is possible with pneumonectomy in carefully selected patients when the expected outcome is balanced against the increased morbidity and mortality rates.
Extended Resection
Patients with STS requiring concurrent en bloc chest wall, pericardial, or diaphragm resection have a significantly shorter median OS than those undergoing only subpleural resection. However, when subjected to multivariant analysis, this difference loses significance, likely secondary to an association with incomplete resection. Therefore, extended resection with en bloc structures may be potentially curative if negative margins can be achieved.
Thoracic Lymph Node Dissection
The role of thoracic lymph node dissection in PM has been studied in mixed populations of patients with epithelial cancer, sarcomatous cancer, germ cell cancer, renal cancer, and melanoma undergoing pulmonary resection. The 5-year OS with lymph node involvement is 0% to 14%, compared with 36% to 60% with no lymph node involvement. Pfannschmidt and colleagues reported on 245 patients who underwent PM along with systematic hilar and mediastinal lymph node dissection. Of these patients, 69 had osteogenic or STS pulmonary metastases. The prevalence of lymph node involvement for patients with sarcoma was 23%. The median survival for the sarcoma group was 47 months for node-negative, 18.3 months for N1, and 22 months for N2 involvement ( P = .036). Hilar and mediastinal lymph node involvement is a poor prognostic indicator and, if identified preoperatively, should be considered a relative contraindication to resection. In the case of occult metastases, further studies of systematic nodal dissection to determine the true prevalence should be considered.
Related posts:
Prognostic Factors and Staging for Soft Tissue Sarcomas: An Update
Margin Status, Local Recurrence, and Survival: Correlation or Causation?
Adjuvant Chemotherapy for Soft Tissue Sarcomas
Isolated Regional Therapy for Advanced Extremity Soft Tissue Sarcomas
Updates on the Management of Gastrointestinal Stromal Tumors
Sarcomas and the Immune System: Implications for Therapeutic Strategies
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
