78: Pulmonary Metastasectomy

The recognition of pulmonary metastases from an extrathoracic primary tumor is a dramatic and emotional change in the care of the cancer patient. The clinical situation immediately changes from potential cure to the tacit acknowledgment of probable incurability. Goals of therapy change from living without evidence of disease to living, and living well, with systemic disease. We believe that pulmonary metastasectomy in carefully selected patients contributes to quality of life and may give the patient extended periods of time without obvious disease.

Many primary tumors metastasize to specific target organs.¹ In the 1930s, it was noted that patients dying of pulmonary metastases frequently failed to exhibit extrapulmonary disease at autopsy.² As a result, several surgeons felt that it would be reasonable to offer surgical resection of these lesions in the hope of prolonging survival. The first reported pulmonary metastasectomy removed a single renal cell metastasis in 1930, and the patient lived for two more decades.³ Interest in this surgical approach was increased with the development of systemic adjuvant chemotherapy, which appeared to increase survival.

The role of surgical resection of metastatic disease, however, is not universally accepted in the nonsurgical community. No randomized trial has been constructed to establish a survival advantage of pulmonary metastasectomy. In fact, the multitude of variables that would have to be included in the eligibility criteria (e.g., number of metastases, cell type, disease-free interval [DFI], cardiorespiratory reserve) makes it unlikely that a randomized trial of this nature will ever be performed. Without such data, however, some authors remain skeptical that surgical resection adds significant benefit.^4,5 Opponents to this approach point out that larger trials claiming improved survival after metastasectomy have been conducted in heterogeneous populations with tumors of mixed histologic types and mixed doubling times. The patients who benefit the most from surgical resection have a small tumor burden and a long doubling time (DFI), and this group may be able to live a long time with their disease even without surgical resection. The skeptics argue that a hypothetical study population that included slow-growing tumors in 40% of the subjects would produce a 30% 5-year survival after surgery, and that same 30% also would still be alive without an operation.

These arguments highlight the heterogeneity of this patient population and emphasize the need to tailor the surgical approach to each individual patient. Patients with dozens of metastases or rapid recurrence after a previous pulmonary metastasectomy will not gain major benefit from surgery. A very elderly patient with a slow-growing metastasis that would require pneumonectomy for resection might be better treated in other ways. These exceptions still leave many patients with pulmonary metastases that can be removed safely with an anticipated low morbidity and mortality.⁶

A large volume of retrospective data is available to substantiate a significant long-term survival with pulmonary metastasectomy. When these data are compared with those of patients with pulmonary metastases who did not undergo metastasectomy, it is likely that pulmonary metastasectomy affords a distinct survival benefit.^7–13 Furthermore, resection has other advantages. It can prevent further growth and consumption of lung tissue (dyspnea), hemoptysis, and chest wall invasion with subsequent pain.

The International Registry of Lung Metastases was established in 1991 to document long-term results and has accrued data on 5206 cases of pulmonary metastasectomy. The distribution of primary malignancies in this data bank was shown to be 43% carcinoma, 42% sarcoma, 7% germ cell tumors, and 6% melanoma. Actuarial 5- and 10-year survival rates after complete metastasectomy were 36% and 26%, respectively. Determinants of improved survival include complete resection, DFI (time from removal of primary tumor to recognition of pulmonary metastases) of 36 months or greater, and single metastasis. These results provide strong evidence that pulmonary metastasectomy is a safe and potentially curative procedure, because it is highly unlikely that a slow tumor doubling time could produce a 26% 10-year survival rate.⁶

Surgical resection of a single pulmonary metastasis has become a widely accepted treatment modality for properly selected patients. Opinions vary widely, however, regarding the utility of surgical resection in the face of multiple metastases, lung and lymph node metastases, or repeat metastases after a previous pulmonary metastasectomy. Each of these topics is explored in this chapter. We also examine the current indications for pulmonary metastasectomy, outcomes of surgery, prognostic indicators by cell type where data are available, and the approach to treatment of extrathoracic pulmonary metastases.

Ehrenhaft et al.¹⁴ published the first “criteria for pulmonary metastasectomy” in 1958. The current indications have remained largely the same but with slight evolution:

1. Control of the primary site

2. No other distant extrapulmonary metastatic disease or, if present, immediate plans to control it with surgery or another treatment modality

3. Pulmonary metastases that are thought to be completely resectable, even if located in both the lungs

4. Adequate cardiopulmonary reserve of the patient

5. A technically feasible operation

Additional criteria for pulmonary metastasectomy have been described and include no other effective treatment except resection, difficulty of differential diagnosis from lung cancer, and symptomatic pulmonary metastases.^15–17 Among patients who do not fulfill the preceding criteria, surgery should proceed only if it will provide ample palliation, such as in situations of bronchial obstruction or distal pulmonary suppuration. The criteria for surgical resection will continue to evolve as we gain a better understanding for cell type-specific tumor biology and as improved systemic treatment becomes available.

Once it is determined that a patient meets the criteria for surgical intervention, the decision to proceed should be made in cooperation with a thoracic surgeon and a medical oncologist. A multidisciplinary approach offers the benefit of coordinating the timing of surgery and systemic treatment both now and in the face of future recurrence.

Preoperative staging and operative planning largely depend on cross-sectional imaging. Radiologic staging is inaccurate in a large proportion of patients, however, and underestimates the burden of disease. In many centers, thorough intraoperative staging by open bimanual palpation is performed in every patient to optimize resection of all metastatic deposits, particularly when more than one lesion is identified on preoperative radiographs.⁶

Helical CT scanners are now widely available. They allow thin-section imaging of the chest and detect approximately 20% to 25% more nodules than conventional CT scanning.^18,19 Large retrospective studies comparing high-resolution helical CT scanning with intraoperative detection of lesions by manual palpation have demonstrated sensitivities of 78% to 82% for CT detection. Approximately 22% of patients would have residual malignant deposits detectable by palpation if CT alone was used to guide resection.^20,21 Helical CT scanning may not be sensitive enough to obviate the need for manual lung palpation if the goal is resection of all detectable metastases. The magnitude of survival benefit with removal of deposits so small that they can only be detected by palpation, however, is uncertain. This is an important issue with the availability of minimally invasive thoracoscopic approaches to pulmonary metastasectomy that preclude manual palpation of the lung.

Helical CT scanning is most likely to miss nodules smaller than 6 mm in diameter. Sensitivity of CT scanning decreases in proportion to the size of the metastatic lesion. Sensitivity is 100% for lesions larger than 10 mm but only 66% for lesions between 6 and 10 mm.²¹ Manual palpation can detect 2-mm nodules at the surface of the lung and 4-mm nodules in the central part of the lobe. Clearly, a CT scan can preclude surgery if the number of nodules prevents a feasible operation.

There is no pathognomonic radiographic feature that distinguishes metastatic disease from primary lung cancer. That said, metastatic nodules often are observed to be well-circumscribed spherical deposits with smooth margins. They are predominantly subpleural in location or located in the outer third of the lung fields, and when multiple nodules are present, the probability of metastatic disease increases significantly. So-called lollipop nodules, which have the appearance of distinct round nodules at the end of a terminal pulmonary artery branch, also raise suspicion for a metastatic process. Primary bronchogenic lung cancer, conversely, often demonstrates irregular borders and associated linear densities.

The presence of a new radiographic nodule in a patient with prior malignancy represents a cancer in more than two-thirds of cases, probably as a function of nodule size (Table 78-1). For patients with a history of extrathoracic malignancy, the probability of a malignant solitary pulmonary nodule ranges from 67% for subcentimeter nodules to 91% for nodules larger than 3 cm. For all lesions smaller than 3 cm, the chance is essentially equal that the nodule represents a primary lung cancer or extrathoracic metastasis. For nodules larger than 3 cm in size, the probability of a primary lung cancer is greater.²²

Many patient and tumor factors have been found to predict improved survival after metastasectomy. Most commonly, these include a long DFI after treatment of the primary tumor, a low number of pulmonary metastases, and favorable cell type.⁶ The actual determinants may be discordant among different reports, likely owing to variance in definitions, measured endpoints, different grouping of cell types, and perhaps different tumor biology.

One positive predictive element ubiquitous to nearly all reports of pulmonary metastasectomy is the poor consequences of incomplete resection. Complete removal of all metastatic deposits is associated with long-term survival (Fig. 78-1).

DFI is regarded by many authors to be important. However, in a review of 32 retrospective reports of pulmonary metastasectomy, 19 reports determined that DFI was a significant prognostic factor, whereas 13 indicated that it was of no consequence.²³ The heterogeneous and small populations represented in these studies make this point difficult to interpret. The International Registry of Lung Metastases was established in 1991 to answer these types of questions. This registry, reported over 20 years ago, has accumulated its data through voluntary reporting by surgeons throughout the world. DFI is an important prognostic marker in this registry. It is important to note that the DFI is considered in terms of likelihood of achieving long-term disease-free survival but is not an absolute indication or contraindication to surgery.

Studies are evenly divided with respect to whether the number of metastatic lesions affects prognosis after pulmonary metastasectomy. Based on 15 reports in the literature, Todd²³ found that the number of metastatic lesions was a consistent factor for sarcomatous metastases, in which fewer than five lesions was associated with a more favorable outcome.²³ For epithelial tumors, the probability of incomplete resection has been a function of the number of pulmonary nodules (Fig. 78-2)²⁴ and the presence of more than one lesion may increase the probability of lung metastases to recur.²⁵ The important issue for the surgeon considering resection of multiple metastases is the feasibility of the operation. If several nodules need to be removed from the same lobe or a central nodule needs to be removed, a lobectomy may be required. As more lung tissue is removed, the perioperative risks increase.

Histology of the tumor itself is remarkable for the extremes of survival benefit. Germ cell tumors are associated with a very good prognosis. The prognosis associated with multiple melanoma metastases, lung primary metastases to the other lung, and pancreatic metastases is so poor that surgery likely will not benefit the patient.²⁶

Two main principles direct the surgical approach for resection of pulmonary metastases: complete resection of malignancy and maximal sparing of normal lung tissue. Complete resection of metastatic disease is not only the goal but also an important predictor of survival in nearly all reported series. All resections are performed conservatively, leaving as much normal parenchyma as possible. For peripheral nodules, this usually means wedge resection. For deeper nodules, segmentectomy may be required. This conservative approach leaves the patient with a greater degree of cardiopulmonary reserve to withstand subsequent treatments, should a recurrence appear, and maintains the patient's quality of life. Mortality rates of pulmonary metastasectomy do not differ from those of resection for lung cancer and range between 0.6% and 2%.^6,27

Bronchoscopy is indicated in patients with centrally located lesions, in patients with symptoms of airway involvement, and in patients with cell types that are prone to endobronchial involvement, such as breast, colon, and renal cell cancers.²⁸

Single-lung anesthesia is used. Complete atelectasis of the lung will ensure adequate palpation of pulmonary parenchyma in open procedures and permits the technique of thoracoscopic exploration and resection.

The choice of incision for unilateral disease is somewhat controversial. Lateral thoracotomy has been the most commonly described approach to pulmonary metastasectomy. This approach offers adequate access to all areas of the hemithorax and permits wedge or anatomic resection under direct vision.

Some authors favor median sternotomy as an approach to unilateral disease. This incision has the advantages of simultaneous examination of both the lungs, identification and treatment of contralateral occult disease, and reduced postoperative discomfort and deficit in pulmonary function. Its disadvantage in pulmonary metastasectomy is difficult exposure of the posterior costovertebral lung fields and the lateral left lower lobe. The bilateral anterior thoracotomy (clamshell incision) provides excellent exposure to the posterior aspect of both lungs, but is complicated by increased postoperative pain.

Inspection of both hemithoraces in treatment of unilateral disease will identify occult tumor deposits in some cases; however, there does not appear to be a survival advantage to this approach compared with lateral thoracotomy.²⁹ Known bilateral lesions are resected via median sternotomy, the clamshell approach, or two-stage bilateral thoracotomy.

Metastatic lesions involving the diaphragm, chest wall, mediastinum, and pericardium are often resectable. Five-year survival after extended resections including chest wall and diaphragm has been reported to be as high as 25%.³⁰

Video-assisted thoracoscopic surgery (VATS) is a less-invasive alternative to thoracotomy or sternotomy. VATS approaches are applicable for both diagnostic purposes and therapeutic metastasectomy, provided that the lesions are located in the peripheral third of the lung. The advantages of thoracoscopic resection include improved postoperative recovery, shorter hospital stay, and decreased long-term morbidity. Disadvantages include the risk of incomplete resection, which possibly could be alleviated by thoracotomy and direct palpation.

Most pulmonary metastases are located in the peripheral third of the lung and are accessible by a thoracoscopic approach.³¹ Generally, the lesion should be small and located in the outer third of the lung, and endobronchial involvement should be absent. Conversion to an open approach is appropriate for lesions located deeper in the lung parenchyma or if all suspicious lesions cannot be identified in a timely manner. To avoid seeding of the intercostal port site, the specimen should be extracted in a sterile specimen bag.

Much of the controversy over the VATS approach to metastasectomy centers on the belief that manual palpation of the deflated lung is essential to identifying all metastatic diseases, a technique precluded by a thoracoscopic approach. Retrospective and prospective studies have concluded that VATS is an inadequate procedure for pulmonary metastasectomy because identification of all metastatic foci is not possible without a complete manual palpation of the entire lung.^32,33 The number of missed lesions on preoperative conventional CT scanning was too high. One prospective trial was closed after 18 metastasectomies begun using a VATS approach and completed with thoracotomy for complete lung palpation, which resulted in 56% failure rate of CT and VATS to detect all lesions. This study has been criticized for the quality and uniformity of CT imaging used, because high-quality CT imaging is vital to success when considering the VATS approach to metastasectomy.

The bigger question concerning manual palpation of the lung and discovery of all occult lesions during the operation is whether this approach has any effect on survival. It is quite possible that the same long-term results can be achieved with repeat metastasectomy by either repeat VATS or thoracotomy, if occult lesions become apparent clinically.

VATS may be a more appropriate operative choice for certain types of tumors, such as colorectal carcinoma, which most often presents with a single metastasis, as confirmed at thoracotomy.³⁴ It has been shown that the incidence of pulmonary recurrence associated with VATS metastasectomy for colorectal cancer is approximately 20%, which is similar to the findings observed with open approaches.³⁴

VATS is an excellent diagnostic approach for differentiating benign lesions, metastases, or a new primary in patients with solitary pulmonary nodules and a history of antecedent cancer. Among patients with no history of cancer, there is a 63% chance that the solitary pulmonary nodule will be malignant. The chance for malignancy is 82% for patients with antecedent lung cancer, with lung cancer being the most common etiology. Patients with a history of extrathoracic malignancy have a 79% probability of malignancy and roughly even odds that the metastasis is from the previous cancer versus a lung cancer.²² Although CT scan quantification of attenuation, shape, size, and texture characteristics of solitary pulmonary nodules shows promise for accurately differentiating benign from malignant lesions, differentiating metastases from primary lung cancers is not yet possible. In our opinion, resection is the most reliable method for diagnosis and is essential to designing the best therapeutic strategy for maximizing survival. VATS resection is well suited for this objective.

Hilar or mediastinal lymph node involvement sometimes accompanies pulmonary metastases. Determinants of nodal involvement, as well as the prognostic and therapeutic implications, are poorly understood. The frequency with which pulmonary metastases can metastasize to regional lymph nodes is not well characterized and is a subject of debate. In an autopsy series, a 33% incidence of mediastinal lymph node metastases was seen in patients with extrapulmonary carcinoma.³⁵ Several retrospective reviews have reported the incidence of hilar or mediastinal lymph node metastases to be 5% to 28.6%.^6,36–38 Lymph node metastases are found more commonly with carcinoma than with sarcoma.

Several retrospective reviews have attempted to clarify the importance of systematic regional lymph node dissection during pulmonary metastasectomy.^37,38 Only a few reports have addressed the prognostic value of lymph node metastases in patients undergoing metastasectomy. In some series, it is reported that the only factor affecting survival is the presence of metastatically involved lymph nodes.³⁷ In a series of nonsarcomatous extrathoracic primary malignancy, lymph node metastases were found in 29% of the 70 patients undergoing complete lymphadenectomy, and tumor recurrence was significantly higher in patients with lymph node metastases. Three-year survival was 38% for those with lymph node metastases versus 69% for those without. The presence of hilar or mediastinal lymph node metastases predicts poor survival in colorectal^39,40 and renal cell metastases.²⁴ Others have shown a trend rather than a statistically significant difference toward improved survival in patients without involvement of mediastinal lymph nodes when all cell types are studied.³⁸ Clearly, the relevance of nodal metastases should be explored within each cell type.

Although the standard accepted resection for primary lung cancer is either complete lymphadenectomy or lymph node sampling, this is not true for the treatment of pulmonary metastases. Routine lymph node dissection is not performed uniformly. Some authors recommend consideration of complete mediastinal lymph node dissection at the time of pulmonary metastasectomy to improve staging and guide treatment.³⁷ We and other authors believe that a pulmonary metastasis will exhibit the same pattern of lymphatic drainage as a primary lung cancer (Fig. 78-3). In other words, pulmonary metastases can metastasize. If a draining node is confirmed to be involved with metastatic disease at the time of surgery, we recommend anatomic resection when possible.^24,39,40 Some authors recommend the same guidelines for assessment as one would use in a patient with a primary lung cancer, including mediastinoscopy and lymph node dissection. We cannot definitively say whether prophylactic regional lymph node dissection at the time of pulmonary metastasectomy contributes to improved survival, although it does allow more accurate oncologic staging and may signal the need for adjuvant therapy.

Many retrospective series on pulmonary metastasectomy report overall 5-year survival rates of 30% to 40%.^6,27 Data from the International Registry of Lung Metastases showed that after complete resection, the 5-year survival was 33% for patients with a DFI of 0 to 11 months after control of the primary malignancy and 45% for those with a DFI greater than 36 months. For single lesions, the 5-year survival was 43%, and it was 27% for four or more lesions. Global overall survival was 36% at 5 years, 26% at 10 years, and 22% at 15 years after complete resection. These data represent the gamut of epithelial, sarcoma, melanoma, and germ cell tumors, for which the benefit of resection will depend on specific tumor histology (Fig. 78-4)⁶. The literature on cell type-specific experience with pulmonary metastasectomy is growing.

Colorectal Cancer

Although several new chemotherapeutic agents have shown promising effects on systemic metastases, no standard chemotherapy has been found to improve survival significantly in patients with colorectal metastases to the lung. Encouraging surgical results have led to the acceptance of aggressive surgical management. The overall 5- and 10-year survival rates are approximately 35% to 45% and 20% to 30%, respectively.^41–45 Important prognostic factors include number of pulmonary metastases, carcinoembryonic antigen level, and regional lymph node involvement.¹⁷

Renal Cell Cancer

Pulmonary resection for renal cell metastases can offer a 20% to 50% 5-year survival.^24,46,47 Larger number and size of nodules, increasing number of lymph node metastases, shorter DFI, and decreased preoperative forced vital capacity are predictive risk factors for death in this population.²⁴

Germ Cell Cancers

The advent of effective chemotherapy has changed the management of pulmonary metastases of germ cell tumors dramatically, and chemotherapy is now the first line of treatment. Surgical resection has been relegated to an adjuvant form of therapy and is reserved for patients who have a complete serologic response to chemotherapy with normalization of serum tumor markers and residual primary lesions. Although surgery may benefit patients who have had a decrease in serum markers shy of complete normalization, the patient with continued elevation of markers is likely better served with a change in chemotherapy. For most patients with germ cell cancers, pulmonary resection is no longer used as primary treatment.

Cisplatin-based chemotherapy is essential as initial treatment for nonseminomatous testicular cancer. Pathologic findings of mature teratoma or necrosis in resected pulmonary metastases indicate good prognosis, with only a 5% to 10% relapse rate. Pulmonary resection is used both diagnostically and therapeutically to determine whether active microscopic disease is still present after normalization of tumor markers and to remove residual mature teratoma. The 5-year survival rate for patients who have these tumors is 68% to 82% after pulmonary metastasectomy.⁴⁸

Sarcoma

Sarcoma has a natural tendency to metastasize to the lungs. Isolated pulmonary sarcoma metastases occur in 23% to 54% of patients with sarcoma.⁴⁹ Pulmonary metastasectomy is an important method of therapy since most sarcomas are relatively chemoresistent. Several retrospective studies have established 3-year survivals of 20% to 54% for patients with surgical resection of sarcoma lung metastases.⁵⁰ Sarcomas are a heterogeneous group with over 50 different histologic subtypes. Among patients with soft-tissue sarcoma that develop pulmonary metastases, the most common histologic findings include malignant fibrous histiocytoma, synovial sarcoma, and leiomyosarcoma.⁵¹

Osteosarcoma

Operative resection is the only potentially curative treatment for patients with thoracic metastases from osteogenic sarcoma. After the primary tumor has been resected, routine adjuvant chemotherapy can improve the disease-free survival and decrease the burden of metastatic pulmonary disease.⁵² The 5-year survival for pulmonary metastases is related to the timing of appearance with respect to the initiation of chemotherapy. Survival when lesions are detected after completion of chemotherapy is significantly better than with the appearance of metastases during chemotherapy.⁵³ Global 5-year survival rates after pulmonary metastasectomy range from 20% to 50%.^23,28

Soft-Tissue Sarcoma

Soft-tissue sarcomas are notoriously resistant to chemotherapy and almost always metastasize solely to the lungs. Surgery remains the only potential curative treatment, and the median survival after diagnosis of pulmonary metastases without surgery is 15 months. Reported 5-year survival rates after pulmonary metastasectomy for soft-tissue sarcoma range from 25% to 35%.^54,55 Factors associated with improved survival among different reports include prolonged DFI, low-grade tumor, young age, slow tumor doubling time, low number of nodules, histology of malignant fibrocystic histiocytoma, and unilateral disease.^54,55

Breast Cancer

Metastatic surgery for breast cancer is highly controversial because of the availability of other effective systemic treatments such as chemotherapy, hormone therapy, and molecular targeting therapy. Dramatic improvements in these treatment options have made pulmonary resection less common. Large retrospective data sets, however, suggest that resection for lung metastases from breast cancer may provide equal or better long-term results than chemotherapy and hormone therapy, with 31% to 50% 5-year survival.^23,56 Postoperative survival is influenced by estrogen receptor status and DFI in some studies.^56,57 We believe that solitary pulmonary nodules in patients with previous breast cancer are best treated with excisional biopsy because differentiation from primary lung cancer is difficult. In general, these patients are worked up in preparation for possible lobectomy, if the tissue diagnosis is consistent with a lung primary. For multiple metastases from breast cancer, medical treatment should be the first-line therapy after establishing a tissue diagnosis.

Gynecologic Cancers

Encouraging results have been reported for pulmonary metastasectomy of uterine cancers. Five-year survival of 47% is reported for squamous cell carcinoma, 33% to 40% for cervical adenocarcinoma, 76% for endometrial adenocarcinoma, and 86% for choriocarcinoma.⁵⁸

Melanoma

Pulmonary metastases from malignant melanoma are associated with poor survival owing to the aggressive behavior of this tumor and its propensity to metastasize systemically to other sites besides the lung. The largest melanoma series, based on 7564 patients, found a 12% incidence of pulmonary metastases in patients with melanoma and an associated 5-year survival rate of 4%. However, of those patients who underwent resection of a solitary pulmonary nodule, 5-year survival was 20%, and patients with two pulmonary nodules undergoing resection fared better than those with three or more nodules. DFI, negative lymph nodes, and treatment with chemotherapy are also associated with a good prognosis.¹²

Head and Neck Cancers

Postoperative survival after resection of pulmonary metastases from head and neck cancer appears to vary by cell type. Five-year survival rates of 34% for squamous cell carcinoma, 64% for glandular tumors, and 84% for adenoid cystic carcinomas are reported.⁴⁸

Multiple pulmonary metastasectomies may be required when there are isolated recurrences in the lung after initial pulmonary metastasectomy, and these can be accomplished in a safe and effective manner.^6,16 After initial pulmonary metastasectomy, a significant number of patients will have recurrence in the lung. In a study of epithelial tumors, after first pulmonary resection, the disease recurred in 68% of patients, with the lung being the first site of recurrence in 38% of patients.²⁵ After curative resection for metastatic soft-tissue sarcoma, 40% to 80% of patients recurred in the lung. For these patients, there is a clear-cut role for repeat pulmonary metastasectomy.⁵⁴

The role of repeat metastasectomy has been evaluated retrospectively. In the 1997 report from the International Registry of Lung Metastases, 53% of the 5206 patients undergoing pulmonary metastasectomy experienced a recurrence.⁶ Five-year survival in these 1042 patients able to undergo a second operation was 44%, and 10-year survival was 29%. Similar 5-year survival rates have been reported by Robert et al.(48%) and Kandioler et al.(50%).^26,59

In a more recent analysis, 82 patients who underwent surgical resection of pulmonary metastases at the Brigham and Women's Hospital (from 1989 to 2004) were analyzed and 31 (38%) were determined to have leiomyosarcoma. Of these, 15 (48%) had repeated pulmonary metastasectomy. There was no difference in disease-free survival between patients with leiomyosarcoma and those with other histologic subtypes of sarcoma, but there was a marked improvement in overall survival (70 vs. 24 months; p = 0.049). In multivariate analysis, the DFI from primary tumor resection to pulmonary metastases and the DFI from pulmonary metastasectomy to second pulmonary recurrence were identified as independent predictors of survival.⁵⁰ The authors concluded that leiomyosarcoma pulmonary metastases behave more indolently than other sarcoma histologic types, and repeat pulmonary resections are warranted.

Survival decreases with the number of repeat resections, and the probability of benefiting from a repeated attempt at thoracic metastasectomy diminishes after each previous operation (Fig. 78-5). For all tissues types, the 5-year survival after the second metastasectomy is 59%; for those undergoing three procedures, it is 38%; and for those undergoing four or more procedures, the survival is 46 months.¹⁶ All deaths are due to recurrent unresectable cancer.

These reports combined suggest that each successful metastasectomy reestablishes the probability for long-term survival by returning the patient to a clinical local control state (Fig. 78-6). Expected survival drops to 8 months, however, after a recurrence can no longer be removed.¹⁶

We and other authors propose that equal results can be obtained by resecting only radiographically visible lesions without an open search for occult disease by manual palpation, provided that the patient undergoes close imaging follow-up with resection of lesions detected on subsequent scans.^16,50 These patients may achieve equal survival if these metastases can be removed in several minimally invasive operations, provided that the risk of further metastatic spread from the remaining pulmonary focus of cancer is relatively low.

There is adequate experience in the literature to conclude that resection of pulmonary metastases can be performed with extremely low mortality and minimal morbidity. Resection of pulmonary metastases improves survival in most malignancies if complete resection is performed in the setting of adequate pulmonary reserve. VATS is a useful alternative to thoracotomy for resection of metastases, provided that the patient is followed closely postoperatively. Repeat metastasectomy should be considered in any patient with recurrent metastases who fits the criteria for operation.

Resection of pulmonary metastasis is generally indicated, provided that the primary tumor is controlled and all resectable disease is removed. Factors such as tumor doubling time and tumor type, among others, have not been shown to affect long-term survival. Recent data support looking for lymph node metastases, especially in colorectal and renal metastases, and if these are found, then pulmonary resection is not performed. Modern day spiral CT scans probably can identify nodules even below 4 mm in size. The need for palpation therefore is questionable and hence VATS resection is feasible. Sequential removal of all lesions identified will probably lead to as good a long-term survival as open palpation and exploration. The tiny lesions that are missed by CT will likely be picked up later on and removed subsequently.