Surgical resection of pulmonary metastases: Benefits, indications, preoperative evaluation, and techniques

INTRODUCTION — Lung metastases from a primary extrapulmonary malignancy are often a manifestation of widespread dissemination; however, some patients have no other evidence of disease [1,2].

Extensive experience with pulmonary metastasectomy in a number of different cancers has suggested that resection may substantially prolong survival and even cure some patients [3]. Based upon these observations, aggressive resection of isolated pulmonary metastases has become a widely accepted treatment for appropriately selected patients. However, acceptance is not universal, and it needs to be proven whether surgical metastasectomy substantially improves survival [4].

The benefits of metastasectomy, selection criteria, preoperative evaluation, and techniques for surgical resection, as endorsed by the Society of Thoracic Surgeons [5], are reviewed. Outcomes according to histology are discussed elsewhere, as are issues specific to resection of soft tissue sarcoma lung metastases. (See "Surgical resection of pulmonary metastases: Outcomes by histology" and "Surgical treatment and other localized therapy for metastatic soft tissue sarcoma".)

The integration of surgery into multidisciplinary therapy for patients with specific malignancies is discussed in the appropriate topic reviews.

BENEFITS OF RESECTION — While pulmonary metastasectomy is a commonly performed operation, belief in its effectiveness is based upon registry data and surgical follow-up studies. Until well-conducted, larger trials are completed, uncertainty will remain about the effectiveness of metastasectomy compared with other forms of treatment (eg, chemotherapy, stereotactic radiotherapy) [4,6].

There is only one published randomized trial examining the benefits of pulmonary metastasectomy. This trial, from Great Britain (PulMicc, NCT01106261) randomly assigned 65 patients with metastatic colorectal cancer to pulmonary metastasectomy or active monitoring [7]. The study was stopped early for low recruitment. Since the original publication, further data from 28 randomized patients have been added. Survival data based on 18 months of additional follow-up are available. Five-year survival was similar for those randomized to surgery compared with no surgery (36.4 versus 29.6 percent), and median survival was also similar (3.78 versus 3.53, respectively) [8]. However, this trial remains underpowered to conclude that no surgery and surgery are comparable but does document the long-term survival of a substantial cohort of nonsurgically treated patients. The only endpoint measured was survival with little data on need for additional therapies and local control. Another major flaw in this study was that molecular prognostic features (eg, BRAF and RAS mutations), and the contribution of systemic chemotherapy was not delineated. PulMicc has provided some additional evidence that a modest survival benefit may be associated with metastasectomy, but probably not as prominent as previous data suggests, perhaps due to roughly 30 percent five-year survival in the nonsurgical arm (a similar value to that reported in a large retrospective series of 1033 adults undergoing stereotactic radiotherapy for extracranial oligometastatic disease, 44 percent involving lung [35 percent five-year overall survival] [9]). The lack of power in this study made it disappointing and highlights the need for well-performed larger trials with surgical randomization.

There are a growing number of studies analyzing survival and long-term relapse-free survival (ie, cure) for patients who undergo pulmonary metastasectomy. Survival analysis has greatly varied among studies, and survival is likely impacted by patient selection. The results from any one study should be interpreted critically until larger randomized controlled trials are conducted. The following retrospective reviews illustrate the benefits of a complete pulmonary metastasectomy:

●A review from the International Registry of Lung Metastases identified 5206 patients with a variety of primary metastatic tumors (carcinomas, sarcomas, germ cell tumors, and melanomas) who underwent a pulmonary metastasectomy [3]. This series included 4572 (88 percent) in whom complete resection was carried out [3]. The overall 5-, 10-, and 15-year survival rates were 36, 26, and 22 percent, respectively. Factors associated with a better prognosis included germ cell etiology, disease-free interval of 36 months, and a single metastatic lesion.

●These outcomes were further corroborated by a 2017 analysis of the Memorial Sloan Kettering Cancer Center database [10]. Over a 25-year period, 539 patients underwent 760 metastasectomies of soft tissue sarcoma with curative intent. Although the median disease-free survival of 6.8 months was short, there was a long tail to the Kaplan-Meier survival curve, 34 percent surviving five years, and 23 percent surviving seven years. (See "Surgical resection of pulmonary metastases: Outcomes by histology", section on 'Soft tissue sarcoma'.)

●A retrospective review of 178 consecutive patients undergoing 256 surgical resections for pulmonary metastasis identified a complete resection in 248 cases [11]. The five-year survival based upon primary disease site and complete resection was the highest for patients with renal cell carcinomas (51.4 percent), followed by colon or rectal cancer carcinoma (50.3 percent), sarcoma (21.7 percent), and melanoma (25 percent).

●Another study of pulmonary metastasectomy is ongoing in Belgium and the Netherlands [12]. However, this prospective nonrandomized phase II multicenter trial is testing the benefit of adding isolated lung perfusion with melphalan at the time of surgical metastasectomy and is not comparing surgical metastasectomy versus systemic chemotherapy, active monitoring, or local ablation. To date, 107 patients have been treated, with no operative mortality and 42 percent grade 3 morbidity. The initial three-year overall survival rate was favorable at 57 percent and the median survival time was 44 months. An update at five years reported 0 percent mortality with a median time for colorectal carcinoma to local progression, median time to progression, and median survival of 31, 14, and 78 months, respectively. In sarcoma patients, local progression was not reached, but median time to progression and median survival were 13 and 39 months, respectively. Five-year disease-free rate and pulmonary progression-free rate were 26 and 44 percent for colorectal carcinoma and 29 and 63 percent for sarcoma [13]. Despite the lack of randomization to resection versus systemic chemotherapy or other local forms of treatment, this study adds additional evidence that surgical resection with or without adjuvant therapy provides an opportunity for long-term recurrence-free survival. Interestingly, the Consensus Statement from the Society of Thoracic Surgeons specifically stated: "Outside of clinical research, isolated lung perfusion is not warranted for management of pulmonary metastases" [5].

●A review from a single institution from 2008 to 2018 identified 281 patients with metastases from a wide range of tumors who had undergone pulmonary metastasectomy for curative intent [14]. Complete resection was achieved in 97.5 percent, and five-year overall survival rate after the first metastasectomy was 47 percent. Primary tumor origin, age, and completeness of the resection were significant independent prognostic factors. Tumor of genitourinary origin and age <66 years were associated with an improved prognosis.

●In a review of 476 patients treated who underwent pulmonary metastasectomy for oligometastatic disease from various primary tumors from 2000 to 2017, overall survival was 88.9 percent at one year and 49.6 percent at five years [15]. Factors associated with improved survival included R0 resection; colorectal primary cancers; and sublobar, nonanatomic resections.

PATIENT SELECTION — The decision to proceed with pulmonary metastasectomy requires a multidisciplinary approach that includes the medical oncologist, radiation oncologist, and thoracic surgeon [5]. The goal is to offer surgery to only those patients who are most likely to benefit, either in terms of prolonging survival or symptom palliation, and to optimize the timing of surgical intervention. Where possible, all cases should be reviewed at a multidisciplinary tumor board prior to scheduling surgical metastasectomy.

Criteria for considering a resection — Most clinicians would accept the criteria listed in the table (table 1) as meeting suitability of pulmonary metastasectomy [16-18].

There are few published guidelines from expert groups that deal with patient selection for pulmonary metastasectomy. Available guidelines include the consensus-based guidelines for resectability of colorectal cancer lung metastases from the National Comprehensive Cancer Network (NCCN) [19] and a consensus document on pulmonary metastasectomy from the Society of Thoracic Surgeons (STS) [5].

For patients who do not meet the criteria (table 1), stereotactic radiotherapy, radiofrequency ablation, or cryoablation offer an alternative to resection for local control. (See "Image-guided ablation of lung tumors" and "Radiation therapy techniques in cancer treatment", section on 'Stereotactic radiation therapy techniques'.)

Contraindications

●The presence of brain metastases is a contraindication to pulmonary resection.

●N2 nodal involvement is generally a contraindication for pulmonary metastasectomy for tumors other than renal cell cancer.

●Inability to completely resect all areas of pulmonary involvement is widely considered to represent a contraindication to pulmonary metastasectomy. Nearly all reports indicate that complete resection of metastatic disease is associated with more favorable outcomes. In the International Registry series, for example, the median survival with complete versus incomplete resection was 35 versus 15 months, and the five-year survival rate was 36 versus 13 percent [3].

●A prior pneumonectomy is a relative contraindication to contralateral resection of pulmonary metastases.

Factors that may influence the decision — Multiple factors influence survival following metastasectomy, although the presence of one or more of these poor prognostic factors does not necessarily constitute an absolute contraindication to metastasectomy [20]. Unfavorable prognostic factors include an inability to completely resect all metastatic disease, a short disease-free interval following treatment of the primary tumor, a large number of pulmonary metastases (the STS consensus group considers pulmonary metastasectomy to be best suited to patients harboring three or fewer metastases [5]) and involved lymph nodes. Tumor histology also influences outcomes. (See 'Lymph node spread' below and "Surgical resection of pulmonary metastases: Outcomes by histology".)

Disease-free interval — Outcomes are more favorable when there has been a longer disease-free interval between primary tumor treatment and presentation of metastatic disease; however, there is no absolute time frame (including synchronous presentation of metastatic disease) that is short enough that metastasectomy would not be considered. However, particularly for a presentation of synchronous metastases, we recommend repeat chest computed tomography (CT) six to eight weeks after recognition of pulmonary metastases to assure that additional target lesions (or too many target lesions) have not appeared. (See 'Timing of surgery' below.)

Although early series yielded conflicting results about the importance of disease-free interval on outcomes [21], data from the International Registry indicate higher five-year survival rates for patients with a disease-free interval of greater than 36 months (45 versus 33 percent for a disease-free interval of less than one year) [3].

Number of lesions — Outcomes are better with fewer metastases. However, for patients with multiple metastases, there is no consensus among thoracic surgeons as to what disease burden represents an insurmountable obstacle. The important issue is the feasibility of resecting all sites of disease, not the absolute number of metastases per se. However, the STS consensus group considers pulmonary metastasectomy to be best suited to patients harboring three or fewer metastases [5].

Data from the International Registry indicate better five-year survival for patients with a single metastatic focus (single focus 43 percent versus 34 and 27 percent for those with two to three, or greater than three metastases, respectively) [3]. This observation holds true both for patients with sarcomas and epithelial tumors [21,22].

The inverse relationship between the number of metastases and survival is likely due to multiple factors. The higher the number of pulmonary metastases, the greater the probability of incomplete resection (figure 1) [21], the likely heavier burden of occult disease diffusely in the lung, and the greater the probability of disease recurrence in the lung [23]. Additionally, if several nodules need to be removed from the same lobe or a more centrally located nodule needs to be removed, a lobectomy may be required rather than wedge resection. As more lung tissue is removed, perioperative risk increases.

A more important issue than the number of metastases is the feasibility of resecting all sites of disease, although the two are related. In a series of 92 patients undergoing pulmonary metastasectomy for renal cell cancer, the chance of a complete resection was over 80 percent for patients with three or fewer nodules identified on preoperative CT [22]. In contrast, if there were more than six nodules on the CT, the chance of a complete resection was no more than 20 percent.

If resection is not feasible at the time of exploration (eg, for technical considerations such as an incomplete pulmonary fissure or proximity to a major vascular structure that was inapparent on the preoperative radiographs), intraoperative ablation using either cryotherapy or radiofrequency ablation (RFA) could be considered. Another alternative is the placement of fiducials for postoperative irradiation. Where available, techniques such as frameless stereotactic radiosurgery provide an opportunity to treat focal areas of the lung to high radiation doses, while preserving adjacent normal tissue [24].

Lymph node spread — Pulmonary metastases may be accompanied by hilar and/or mediastinal lymph node involvement (figure 2) [25]. Lymph node involvement is an important negative prognostic factor in patients undergoing metastasectomy, regardless of histology [22,26-32]. Sarcomas and most cancers that spread to the lungs do so through the vascular system, and the metastases appear limited to the parenchyma. However, other malignancies, notably renal cell carcinoma, colon cancer, breast cancer, and melanoma, have demonstrated the ability to metastasize to the lung, and then to metastasize to the ipsilateral intrapulmonary or hilar (N1) nodes. This is believed (although not proven) to represent spread through the lymphatic system, in a manner similar to primary lung cancer, and with the capability to spread further to the ipsilateral mediastinal (N2) and contralateral mediastinal (N3) nodes. The frequency of lymphatic involvement in patients with pulmonary metastases has not been fully characterized, in part because a systematic lymph node dissection has not been considered a standard approach in this setting. In retrospective reviews, unsuspected metastatic disease in hilar or mediastinal lymph nodes was revealed in 5 to 33 percent of cases, depending on histology (more common in patients with carcinoma than sarcoma) [3,26-28,33-36].

Lymph node involvement is an important negative prognostic factor for pulmonary metastasectomy, regardless of histology. Surgical or endoscopic staging of the mediastinal and hilar lymph nodes prior to pulmonary metastasectomy provides both diagnostic and prognostic information. We agree with guidelines from the STS that recommend regional nodal sampling for evaluation of nodal metastases prior to attempted pulmonary metastasectomy in the presence of suspicious N1 or N2 nodes on preoperative staging imaging and for clinically node-negative tumors associated with nodal disease spread (ie, renal cell, breast, colorectal carcinoma, and melanoma) [5]. (See 'Operative staging' below and 'Lymphadenectomy' below.)

In general, N3 nodal involvement represents a contraindication to potentially curative metastasectomy for any histology. Outcomes are generally poorer for those with N2 rather than just intrapulmonary or hilar nodal involvement, and we consider N2 involvement to represent a contraindication to pulmonary metastasectomy for tumors other than renal cell cancer [27,28,30,32,35]. For patients with renal cell cancer and N2 nodal metastases, the decision to pursue pulmonary metastasectomy along with surgical lymphadenectomy must be individualized. The prognosis is also worse compared with those who have intrapulmonary or hilar nodes, but not all patients with N2 disease and renal cell cancer are destined to recur [22,37,38]. As an example, in one report of 122 patients undergoing systematic lymph node dissection prior to lung metastasectomy for renal cell cancer, median survival was shorter for those with mediastinal as compared with hilar-only nodal metastases (74 versus 32 months), but long-term survival could be achieved in some cases [38]. (See "Surgical resection of pulmonary metastases: Outcomes by histology".)

Lymphadenectomy is discussed in further detail below. (See 'Lymphadenectomy' below.)

Repeat metastasectomy — For both epithelial cancers [23] and sarcomas [39], the most frequent site of relapse following pulmonary resection is in the lung. In general, sequential resections can be carried out safely to treat isolated recurrences [3,17]. Repeat metastasectomy should be strongly considered for slow-growing or few metastases, provided the patient can tolerate repeat surgeries and will have adequate cardiopulmonary reserve [40].

The role of repeat metastasectomy has been evaluated retrospectively [3,41,42]. The largest series, the International Registry of Lung Metastases, found that 53 percent of the 5206 patients undergoing resection of pulmonary metastases relapsed in the lung [3]. Among the 1042 patients who underwent repeat metastasectomy, the 5- and 10-year survival rates were 44 and 29 percent, respectively. In contrast, the median survival among patients who were not resected following relapse in one series was eight months [17].

These reports suggest that repeat surgery can reestablish control in the chest in carefully selected patients. The likelihood of permanent disease control decreases with multiple sequential resections (figure 3) [17]. In a series of 54 patients from one institution who underwent multiple resections for pulmonary metastases, the percentage of patients achieving durable local control of disease after the second, third, fourth, and fifth procedures was 27, 19, 8, and 0 percent, respectively [17]. When local control in the chest is lost, survival becomes dismal, with an average life expectancy of a few months (figure 4).

The ability to treat pulmonary metastatic disease with repeated resections provides support for the concept that surgery can be restricted to radiographically visible lesions without an open search using manual palpation. Patients managed this way require close follow-up, with resection of lesions subsequently detected by imaging [17,42]. (See 'Surgery' below.)

PREOPERATIVE EVALUATION — Appropriate patient selection is key to successful surgical resection of pulmonary metastases. Some centers routinely evaluate the mediastinal lymph nodes only in patients with suspicious findings on computed tomography (CT), while others advocate routine systematic preoperative staging of the mediastinum in all patients. Prior to planned resection, we suggest evaluation of the mediastinal lymph nodes as would be done for a patient with a primary lung cancer. (See 'Lymph node spread' above and 'Lymphadenectomy' below.)

In addition to thoracic CT, several other studies and a diagnostic biopsy may be indicated. We generally recommend an 18-F fluorodeoxyglucose (FDG)-positron emission tomography (PET) scan prior to pulmonary metastasectomy to optimize the selection of candidates for resection. All extrathoracic areas of suspicious FDG uptake should be vigorously investigated prior to surgery, and surgery in the chest should not be performed unless all known deposits of disease are being treated.

The presence of pathologic mediastinal adenopathy on PET (or CT) should prompt a biopsy via mediastinoscopy, or endobronchial ultrasound-guided fine needle aspiration cytology. (See 'Computed tomography' below and 'PET scans' below and "Endobronchial ultrasound: Indications, contraindications, and complications".)

Computed tomography — Accurate preoperative assessment and operative planning are dependent upon cross-sectional imaging. Thin-section imaging of the chest using high-resolution helical CT is the preferred modality, detecting approximately 20 to 25 percent more nodules than conventional CT, and the reliable detection of nodules as small as 2 to 3 mm [43,44]. However, this improved sensitivity comes at the expense of specificity [45].

Differential diagnosis — A new pulmonary lesion in a patient with a known malignancy may represent a metastasis, a second primary lung cancer (particularly if the patient is a smoker), or a benign lesion.

There are no pathognomonic radiographic features that distinguish metastatic disease from a new primary lung cancer or from benign processes. Metastatic nodules are typically well-circumscribed, spherical deposits with smooth margins and are predominantly subpleural or located in the outer third of the lung fields. In contrast, primary lung cancers are usual single, often have irregular borders and associated linear densities, and more often are located centrally. (See "Diagnostic evaluation of the incidental pulmonary nodule".)

When multiple nodules are present, the probability of metastatic disease increases significantly. However, an exception to this general rule is adenocarcinoma in situ (aka minimally invasive adenocarcinoma or adenocarcinoma of the lepidic spectrum, formerly called primary bronchioloalveolar carcinomas), which may present with multifocal pulmonary nodules, and ground glass opacification, which may be seen with severe acute and chronic nonmalignant pulmonary disease. (See "High resolution computed tomography of the lungs", section on 'Ground-glass opacification' and "Pathology of lung malignancies", section on 'Adenocarcinoma'.)

A new solitary pulmonary nodule appearing in a patient with a prior cancer will be malignant in a majority of cases. In a series of 1104 cases undergoing resection at a single institution, among patients without a history of antecedent malignancy, 63 percent of resected solitary nodules were malignant, while in patients with a history of lung or extrapulmonary cancer, approximately 80 percent were malignant [46].

Nodule size also influenced the likelihood that a lesion was malignant and that it represented a primary lung cancer rather than a metastasis [46]. For patients with a history of prior extrathoracic malignancy, the probability that a solitary pulmonary nodule was malignant ranged from 67 percent for nodules ≤1 cm in diameter to 91 percent for nodules >3 cm. For lesions ≤3 cm, there was an equal likelihood that the nodule represented a primary lung cancer or a metastasis; lung cancer was more common than metastasis if the nodule was >3 cm. (See "Diagnostic evaluation of the incidental pulmonary nodule".)

Although assessment of the clinical scenario and a lesion's radiographic characteristics by CT may provide clues as to whether an individual lesion is benign or malignant, it is not possible to reliably distinguish a metastasis from a primary lung cancer. Resection of the nodule is the most reliable method for establishing the diagnosis and is essential to design the most adequate therapeutic strategy to maximize survival.

PET scans — Although clinical practice varies [47], we recommend a PET scan prior to metastasectomy to optimize patient selection for pulmonary metastasectomy. PET is not necessarily more sensitive than high-resolution CT for detection of pulmonary metastases [48]. Like CT, PET has limited sensitivity for lesions <1 cm in size [49], and sensitivity is not improved by integrated PET/CT.

The results of the PET scan also cannot be used reliably to classify CT-discovered thoracic lesions as benign or malignant [50]. Furthermore, a negative PET scan of the chest should not influence the decision to pursue resection. If enlarging lung nodules (particularly if they have the characteristic CT appearance of metastases) are present without other recognized deposits of disease, we still favor surgical resection if possible for both diagnostic and potentially therapeutic purposes, even if the PET scan is negative. (See "Diagnostic evaluation of the incidental pulmonary nodule", section on 'Positron emission tomography/computed tomography' and "Diagnostic evaluation of the incidental pulmonary nodule", section on 'Introduction'.)

The main value of PET is its high level of sensitivity for detecting extrathoracic disease. Metastasectomy should not be done unless all known disease deposits are being treated (except if the patient is enrolled in a protocol such as a vaccine trial, or for symptomatic lesions that cannot be treated by other means). (See 'Criteria for considering a resection' above.)

In one of the only studies to specifically address the benefit of PET in this setting, 86 patients with proven or suspected lung metastases, deemed potentially resectable after conventional CT scan, underwent preoperative PET [51]. In 19 cases (21 percent), lung surgery was excluded based on the PET results, due to the finding of extrapulmonary metastases (11 cases), an unexpected primary site recurrence (2 cases), mediastinal lymphadenopathy (2 cases), or benign disease (4 cases).

The benefit of PET for patients with soft tissue sarcoma metastases (where extrathoracic metastases are rare) is less certain. (See "Surgical treatment and other localized therapy for metastatic soft tissue sarcoma".)

Positive extrathoracic or mediastinal uptake on a PET scan is not sufficient evidence to exclude a patient from surgery for lung metastases. All suspicious extrathoracic areas should be vigorously investigated (usually with a diagnostic biopsy) prior to surgery.

Brain imaging — We routinely perform brain imaging prior to resection of apparently isolated pulmonary metastases for patients who have tumors that frequently metastasize to the brain (eg, breast cancer, melanoma). The presence of brain metastases can be considered a contraindication to pulmonary resection. (See 'Contraindications' above.)

Need for preoperative tissue diagnosis — Ultimately, the diagnosis of pulmonary metastases is made by tissue sampling. In patients with highly suspicious lesions that have a characteristic CT appearance (eg, round, peripheral, multiple), definitive tissue diagnosis is often made after surgical resection of the metastases. However, in many cases, preoperative CT-guided fine needle aspiration (FNA) biopsy is a useful means to obtain tissue less invasively, particularly if the diagnosis of metastatic disease is in question, the patient is a poor or borderline operative candidate, or the patient has a primary tumor (eg, testicular germ cell tumor, lymphoma) for which surgery may not be necessary.

Particularly for lesions less than 1 cm in size, video-assisted thoracoscopic surgery (VATS) may be useful to establish a diagnosis. (See 'VATS procedure' below.)

Bronchoscopy (with or without endobronchial ultrasound) is indicated as part of the evaluation in cases of centrally located lesions identified on CT, in patients with symptoms of airway involvement, and for cell types that are prone to endobronchial involvement such as breast, colon, and renal cell cancer [52]. Bronchoscopy should be performed preoperatively when positive findings would contraindicate operative intervention. For patients who require general anesthesia, bronchoscopy is performed in the operating room as a component of a staging procedure prior to metastasectomy. (See "Diagnostic evaluation of the incidental pulmonary nodule" and "Image-guided bronchoscopy for biopsy of peripheral pulmonary lesions".)

TIMING OF SURGERY — For patients with pulmonary metastases, particularly if they have a short disease-free interval or a synchronous presentation, we typically delay open surgical intervention for six to eight weeks to evaluate the natural history of untreated disease through repeat computed tomography (CT) scanning, although the specific interval (ie, 6 to 8 versus 12 weeks) has not been rigorously studied. In this manner, patients who rapidly develop extrapulmonary metastases can be spared the morbidity of unnecessary thoracic surgery. This also ensures that a small number of mature (eg, large and fairly uniform in size) target lesions have not evolved into innumerable lesions with a constellation of sizes, which may change the surgical approach.

Patients with one or more pulmonary metastases are likely to develop additional pulmonary lesions over time. Following potentially curative metastasectomy, additional lesions are identified in 50 percent of cases due to the subsequent growth of previously undetectable lesions. Smaller lesions have a shorter doubling time and, therefore, a faster growth rate than larger lesions [53]. New lesions are recognized when these tumor deposits grow and become recognizable as different from surrounding lung tissue [3]. Delaying intervention once a pulmonary metastasis is first identified may allow initially occult metastases to become clinically apparent, permitting a more complete resection and better long-term outcomes. This was illustrated in a retrospective review of 68 patients who underwent pulmonary metastasectomy [54]. On multivariate analysis, the only independently significant prognostic factor for survival was longer interval (greater than twelve weeks) from detection of pulmonary metastases to resection.

Timing for metastasectomy, however, may depend upon the anticipated surgical approach (open versus minimally invasive). For those with a deep nodule or numerous nodules of varying sizes that might require open thoracotomy, which is associated with longer recovery and more morbidity, allowing the full extent of disease to be revealed by delaying open surgery may be beneficial for the patient. In this situation, we delay open surgical intervention for six to eight weeks in order to evaluate the natural history of untreated disease through repeat CT scanning. On the other hand, we may choose not to delay a VATS metastasectomy for an isolated peripheral nodule in a favorable location because recovery time and surgical risk are minimal, and recurrent disease is technically easier to resect with a repeat VATS approach because of fewer adhesions. The finding of equivalent long-term survival rates for recurrent disease following VATS or open metastasectomy implies that there may be little clinical significance to micronodules missed with initial VATS that are resected at a later date [55,56]. (See 'VATS procedure' below.)

SURGERY — The goal of curative resection of pulmonary metastases is complete identification and removal of all foci of malignancy with preservation of a maximal amount of normal lung tissue. If mediastinal adenopathy is suspected or detected by computed tomography (CT) of the chest or positron emission tomography-computed tomography (PET-CT), then endobronchial ultrasound-guided biopsy or cervical mediastinoscopy may be indicated for pathologic mediastinal nodal staging. (See 'Lymphadenectomy' below and "Surgical evaluation of mediastinal lymphadenopathy".)

Operative staging — For patients with suspicious N1 or N2 nodes on preoperative staging imaging, mediastinoscopy should be performed prior to the planned metastasectomy. Suspicious lymph nodes during mediastinoscopy can be sampled and sent for frozen section [34,57], if endobronchial ultrasound-guided biopsy has not been performed. (See "Procedures for tissue biopsy in patients with suspected non-small cell lung cancer" and "Surgical evaluation of mediastinal lymphadenopathy".)

The need for routine systematic lymph node sampling during pulmonary metastasectomy is controversial. We agree with guidelines from the Society of Thoracic Surgeons (STS) that recommend regional nodal sampling for evaluation of nodal metastases prior to attempted pulmonary metastasectomy in the presence of suspicious N1 or N2 nodes on preoperative staging imaging [5]; we also perform this for clinically node-negative carcinomas associated with a higher frequency of nodal disease spread (ie, renal cell, colorectal carcinoma, breast cancer, and melanoma). Most clinicians consider the presence of confirmed mediastinal (N2) nodal metastases to represent a contraindication to metastasectomy because of the poor long-term outcome, at least for tumors other than renal cell cancer (figure 2). In contrast, the presence of hilar or intrapulmonary nodal metastases is not necessarily a contraindication to metastasectomy (especially for renal cell cancer) as long as anatomic resection is possible. (See 'Lymph node spread' above and 'Lymphadenectomy' below.)

Intraoperative pulmonary palpation may also be performed as part of the overall staging for those undergoing open thoracotomy [58]. (See 'Open thoracotomy' below.)

Surgical approach — Many standard lung resections, characterized in name by the amount of lung tissue removed (eg, wedge resection, segmentectomy, lobectomy), can be used to resect pulmonary metastases using an open thoracotomy incision (anterior thoracotomy, posterior thoracotomy) or minimally invasive techniques [59]. Pneumonectomy should be discouraged as it is uncommonly curative in this setting. Whether to choose VATS or an open thoracotomy approach depends upon characteristics of the metastases, including their location, number, and size as well as lesion stability as judged on CT scan. A conservative approach, removing the least amount of lung tissue needed to render the patient free of disease, results in greater cardiopulmonary reserve to withstand subsequent treatments should there be a recurrence.

VATS versus open surgery — Open thoracotomy was traditionally the standard for resection of pulmonary metastases to allow for bimanual palpation of the lungs. However, with improved imaging techniques that can detect smaller lesions with greater accuracy, VATS is gaining popularity. In our view (and that of the STS [5]), minimally invasive approaches are preferred.

Retrospective studies have demonstrated equivalent overall survival between VATS versus thoracotomy, with the former associated with improved perioperative outcomes. There are currently no randomized controlled trials comparing these approaches [60]. (See "Overview of pulmonary resection", section on 'Lung anatomy and types of lung resection'.)

Margins — Obtaining tumor-free adequate margins with a complete resection is important for reducing the risk of recurrence. There is no clear recommendation regarding what is considered an adequate margin for metastasectomies. Various recommendations have been made over the years, with some citing a wedge-shaped resection with clear margins 0.5 to 1 cm in all directions [52] and others recommending a margin of at least 1 cm and tumor/margin ratios less than 1.7 for wedge resections [61].

An adequate margin is usually easily accomplished if the metastasis is peripheral (eg, outer third of lung field as seen on chest CT). For more centrally located lesions, an adequate margin may not be achieved, because excessive tissue tension will prevent the closure of the remaining lung tissue. Thus, central lesions are often better approached with either segmentectomy or lobectomy, and open thoracotomy may be preferred, though lobectomy can also be performed using VATS.

Wedge resection is often sufficient for peripheral nodules, but segmentectomy, lobectomy, or even pneumonectomy may be required for lesions at the hilum or in the central part of the lobe. The mortality of pulmonary surgery is related to the amount of lung tissue removed, with mortality rates of 4 to 10 percent for pneumonectomy, and consistently less than 1 percent following wedge resection. These mortality rates do not differ from those associated with resection for lung adenocarcinoma [3].

In general, a wedge resection by VATS is most appropriate for stable, peripheral metastases, each of which is no greater than 3 cm in size. There should be a small number of nodules; a single metastasis is ideal. Numerous target lesions may preclude a VATS approach, although there is no consensus among thoracic surgical oncologists or sarcoma specialists as to what disease burden represents an insurmountable obstacle.

VATS procedure — Minimally invasive approaches to thoracic surgery have become increasingly popular and can be described as "thoracoscopic" or "video-assisted." Thoracoscopic surgery involves multiple ports, but none larger than a fingerbreadth, while video-assisted thoracoscopic surgery often incorporates a minithoracotomy incision no longer than the length of a finger (10 cm). Minimally invasive techniques are often collectively referred to as "VATS" (video-assisted thoracoscopic surgery), although the literature is variable, and terminology is not standardized [62]. We will use the term "VATS" to refer to techniques not using standard thoracic incisions.

The first series using VATS to perform metastasectomy was reported in 1993 [63]. Thoracoscopy involves the passage of a videoscope through the chest wall, giving the surgeon a "window" for direct visualization into the chest. Additional ports are placed through the chest wall to allow for various instruments. VATS is less invasive than open thoracotomy; however, like thoracotomy, it must be performed in the operating room under general anesthesia with single-lung ventilation. (See "Medical thoracoscopy (pleuroscopy): Equipment, procedure, and complications".)

Because of fewer and smaller incisions, VATS is associated with less pain, faster postoperative recovery, shortened hospital stay, and decreased long-term morbidity (table 2). Intrathoracic postoperative adhesions are milder with VATS compared with open thoracotomy. VATS has the advantage of shortened recuperation time and decreased blood loss in patients who require repeat resection after recurrence [64].

VATS can be used to resect most metastatic lesions as the majority are located in the peripheral one-third of the lung [65]. Conversion from a VATS approach to an open thoracotomy is appropriate for lesions located more centrally or whenever the completeness of resection is in doubt. Intrathoracic recurrence rates of clinically significant disease are reported to be between 18 and 49 percent for VATS, similar to that reported with open thoracotomy (16 to 53 percent) [64].

To avoid intrathoracic and port site seeding, the resected specimen should be placed intact into a sterile specimen bag before being delivered through the port site. It is important to examine the cut surface of the resected specimen to confirm that the resection margin is free of tumor; frozen-section histologic analysis should be undertaken when indicated.

As discussed above, there is no consensus regarding a preference for open thoracotomy over a VATS procedure. Some argue that thoracoscopic resection is a valid approach only for patients with a solitary pulmonary nodule on preoperative chest CT, where the risk of additional disease is small [66-69]. On the other hand, proponents of VATS argue that while open surgical exploration identifies radiographically occult nodules, these can be resected with VATS in the future, once they become detectable on surveillance CT scans, without adversely affecting survival. A review of the European Society of Thoracic Surgeons database from 2007 to 2019 identified 8868 patients who underwent a pulmonary metastasectomy (63.5 percent thoracotomy and 36.5 percent VATS) [70]. Notably, the rate of VATS procedure increased from 15 percent in 2007 to 58 percent in 2018. So, although a clear consensus has not been reached regarding preferably performing VATS versus thoracotomy for pulmonary metastasectomy, VATS is slowly gaining popularity. (See 'Repeat metastasectomy' above.)

However, a VATS option that offers access to both sides of the chest cavity as well as manual palpation involves the use of a subxiphoid port. While reportedly used for bilateral pulmonary resection, the learning curve is steeper compared with use of a standard transthoracic incision [59,71-73]. (See "Overview of minimally invasive thoracic surgery", section on 'Incisions'.)

There are no randomized trials comparing outcomes between open and thoracoscopic resection and few retrospective comparisons. Data suggest that patients do at least as well after VATS as after thoracotomy and imply that there is little if any clinical significance to skipped nodules. However, because these favorable outcomes from VATS were obtained predominantly in populations with a limited disease extent, many surgeons restrict VATS to patients with a single or very few pulmonary metastases. Available studies are inherently biased because patients with single metastases are more often offered VATS.

Specific data illustrating these points include:

●In a retrospective cohort study of 186 patients undergoing pulmonary metastasectomy (36 with VATS and 135 via thoracotomy), five-year overall survival (70 versus 59 percent) and recurrence-free survival (67 versus 51 percent) were not worse after VATS [55]. However, the VATS group more often had single metastases (56 versus 21 percent) and unilateral disease (94 versus 67 percent).

●Another study attempted to minimize patient selection bias by comparing outcomes from both procedures in a relatively homogeneous population of patients with two or fewer sarcoma metastases per lung field [56]. When the 31 patients treated with VATS were compared to 29 patients undergoing open thoracotomy, overall survival and ipsilateral recurrence rates were similar in both groups. In this study, again, VATS patients more often had single lesions (58 versus 38 percent). Comparative rates of perioperative morbidity and mortality were not addressed in this series.

●A retrospective review from Switzerland identified 251 patients who underwent pulmonary metastasectomy from 2009 to 2017 (63 via VATS and 188 via thoracotomy) [74]. Solitary metastases were present in 54 of 63 (85.7 percent) VATS patients. The recurrence rate was lower for VATS patients compared with thoracotomy patients (12.7 versus 22.3 percent), but the observed difference was not significant. There was no difference in median overall survival between the two groups.

Open thoracotomy — The location of the thoracotomy incision depends on the anatomic location of the metastases (anterior and posterolateral approaches are most common), and whether the disease is unilateral or bilateral. Resection of pulmonary metastases by a thoracotomy incision uses one-lung anesthesia for complete atelectasis of the nonventilated lung. Open thoracotomy exposes the chest cavity, facilitating bimanual palpation of the pulmonary parenchyma, and resection of visible and palpable metastases.

A bilateral approach is indicated only for those patients who present radiographically with bilateral metastases. For patients who appear to have unilateral disease, inspection of both hemithoraces frequently identifies occult tumor deposits. However, bilateral exploration and resection has not produced a survival advantage compared with a unilateral approach [75]. Further, recurrence in the ipsilateral lung was as common as recurrence in the contralateral lung in patients who underwent bilateral exploration with complete resection to treat unilateral disease.

The open approach options to patients with known bilateral lesions are median sternotomy, bilateral anterior thoracotomies (clamshell incision), or bilateral sequential thoracotomies (table 2).

●Sternotomy (figure 5) allows the simultaneous examination of both lungs and identification and treatment of contralateral occult disease. The main advantages of this approach are reduced postoperative pain and less impairment of pulmonary function (table 2). Disadvantages include more difficult exposure of the costovertebral lung fields and limited visualization of the left lower lobe because of the heart. A thoracotomy extension may improve exposure (figure 6). The sternal approach may also not be feasible in a patient who has had a previous sternotomy (eg, heart bypass). Longitudinal sternal transection carries a risk of a sternal wound infection; however, the incidence is low. Sternal wound infection (as with any wound infection) can delay recovery and initiation of adjuvant therapy. (See "Surgical management of sternal wound complications".)

●The clamshell incision (bilateral anterior thoracotomies) (figure 7) provides excellent exposure to the posterior aspect of both lungs but is associated with the most postoperative pain (table 2). Because the sternum is divided horizontally, there is also a risk for sternal wound infection, and because both the internal mammary arteries are divided, the risk may be even greater. (See "Early noncardiac complications of coronary artery bypass graft surgery", section on 'Sternal wound infection and mediastinitis'.)

●Bilateral sequential thoracotomies are a good approach when exposure of the posterior aspects of both lungs and systematic hilar lymph node dissection is needed. Lateral thoracotomy incisions can be painful, and a recuperation interval of six weeks is typically required (table 2).

Role for extended resection — Metastatic lesions extending beyond the lung and involving the diaphragm, chest wall, mediastinum, or pericardium are potentially resectable en bloc, provided negative margins can be achieved. In a systematic review, available studies included between 9 and 42 patients with five-year survival rates that ranged from 10 to 41 percent [76]. A later retrospective review of 1027 patients who underwent lung metastasectomy included 29 patients who had extended pulmonary resections for epithelial (62.1 percent) and sarcomatous (20.7 percent) tumors [77]. Procedures included three resections of the chest wall, one azygos, one diaphragm, four vascular resections/reconstructions, six sleeve resections, and 14 pneumonectomies. Perioperative morbidity (30 day) and mortality were 38 (11 of 29) and 0 percent, respectively. Only one patient had a major complication due to a bronchopleural fistula. Overall survival following a complete extended metastasectomy at two years was 66 percent, at five years was 42 percent, and at 10 years was 36 percent.

Lymphadenectomy — The need for routine systematic lymph node sampling during pulmonary metastasectomy is controversial. We agree with guidelines from the STS that recommend regional nodal sampling for evaluation of nodal metastases prior to attempted pulmonary metastasectomy in the presence of suspicious N1 or N2 nodes on preoperative staging imaging and for clinically node-negative carcinomas associated with nodal disease spread (ie, renal cell, breast, colorectal carcinoma, and melanoma) [5].

As noted above, contralateral mediastinal nodal involvement represents a contraindication to potentially curative metastasectomy in a unilaterally affected lung for any histology. Outcomes are generally poorer in those with N2 (ipsilateral mediastinal) rather than just intrapulmonary or hilar nodal involvement, and we consider this to represent a contraindication to pulmonary metastasectomy for tumors other than renal cell cancer. For patients with renal cell cancer and ipsilateral mediastinal (N2) nodal metastases, the decision to pursue pulmonary metastasectomy along with surgical lymphadenectomy must be individualized. (See 'Lymph node spread' above.)

In general, there are few long-term survivors among those with N2 disease, leading many to conclude that documentation of mediastinal nodal involvement represents a contraindication to potentially curative pulmonary metastasectomy, regardless of histology [27,28,30,32,35]. However, although there are no randomized trials in which patients either did or did not undergo mediastinal node dissection prior to pulmonary metastasectomy, successful removal of involved nodes can produce a complete resection with negative margins, which may contribute to the success of the surgery. Leaving an involved node is equivalent to having a positive margin. There are at least some data that suggest that long-term survival is possible for some patients, especially those with renal cell cancer with mediastinal (N2) nodal disease, although outcomes are worse than if nodal disease is limited to the hilar and intrapulmonary (N1) nodes (figure 2). The bias at the authors' institutions has been to treat such malignancies according to their biologic behavior, which includes a willingness to diagnose N2 and N3 disease by endobronchial ultrasound-guided biopsy, endoscopic ultrasound (EUS), or cervical mediastinoscopy. It also includes a surgical plan to resect these metastases with anatomic resections to achieve a local lymphadenectomy in the surgical specimen. Removal of perihilar nodes can frequently be accomplished by thoracoscopic enucleation, while removal of nodes between the vessels and segmental bronchi often requires segmentectomy or lobectomy. We also perform a mediastinal lymphadenectomy in these cases.

Documentation of N2 nodal diseases may make nonsurgical treatment alternatives (such as chemotherapy, radiation, or radiofrequency ablation) more attractive because of the lower perceived benefit of surgery in this situation. Although some have suggested that knowing the status of the regional nodes can guide adjuvant radiotherapy or chemotherapy [27], the presence of nodal metastases does not change medical management for the vast majority of tumor types metastasizing to the lung (sarcoma, colorectal cancer, breast cancer, germ cell tumor). (See "Surgical resection of pulmonary metastases: Outcomes by histology".)

Follow-up surveillance — The primary goal of surveillance after pulmonary metastasectomy is to detect a recurrence while salvage therapy is still an option. Few studies have specifically addressed surveillance to detect recurrence after pulmonary metastasectomy, and available guidelines do not address this issue [5,11,78]. In the review of the International Registry of Lung Metastases, the median time to recurrence was 10 months [3].

For most patients, following initial complete resection of metastases, we suggest regular radiographic surveillance with chest CT every four months. However, close surveillance is only appropriate for patients who are candidates for further surgical intervention in the event of recurrent oligometastatic disease. In some cases (eg, resected metastatic colorectal cancer), histology-specific guidelines are available, and we follow these guidelines (table 3). (See "Post-treatment surveillance after colorectal cancer treatment", section on 'Resected stage IV disease'.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Colorectal cancer" and "Society guideline links: Colorectal surgery for cancer" and "Society guideline links: Diagnosis and management of lung cancer" and "Society guideline links: Soft tissue sarcoma".)

SUMMARY AND RECOMMENDATIONS

●Benefits of resection – Aggressive surgical resection of lung metastases in appropriately selected patients may offer a chance for extended disease-free survival that may not be possible with systemic therapy. Retrospective experience has shown that the overall actuarial 5- and 10-year survival rates after complete metastasectomy are approximately 36 and 26 percent, respectively. (See 'Benefits of resection' above.)

●Patient selection

•For appropriately selected patients, we suggest surgical resection of lung metastases rather than palliative systemic therapy (Grade 2C). Optimal outcomes are highly dependent on appropriate patient selection. Where possible, all cases should be reviewed at a multidisciplinary tumor board prior to scheduling surgical metastasectomy. (See 'Patient selection' above.)

•There are no widely accepted published guidelines as to appropriate patient selection for pulmonary metastasectomy. We suggest the general selection criteria listed in the table (table 1) (see 'Criteria for considering a resection' above):

•Outcomes are more favorable the longer the disease-free interval between primary tumor treatment and presentation of metastatic disease, although there is no absolute time frame (including synchronous presentation) that is short enough that metastasectomy would not be considered. Particularly for synchronous metastatic disease, a delay of two to three months in metastasectomy is reasonable to allow the natural history of the disease to become manifest. (See 'Factors that may influence the decision' above.)

•Outcomes are better with fewer metastases. However, there is no consensus as to what disease burden represents an insurmountable obstacle. The more important issue is the feasibility of resecting all sites of disease rather than absolute number. However, the Society of Thoracic Surgeons (STS) consensus group considers pulmonary metastasectomy to be best suited to patients harboring three or fewer metastases. (See 'Number of lesions' above.)

●Preoperative evaluation

•Prior to planned resection, contrast-enhanced, high-resolution computed tomography (CT) of the chest should be performed. (See 'Preoperative evaluation' above and 'Lymph node spread' above.)

•For patients without evidence of extrapulmonary disease on chest CT, an 18-F fluorodeoxyglucose-positron emission tomography (FDG-PET) scan is indicated. Positive extrathoracic or mediastinal uptake on a PET scan is not sufficient evidence to exclude a patient from surgery for lung metastases. Suspicious areas should be vigorously investigated (usually with a diagnostic biopsy) prior to metastasectomy. (See 'Preoperative evaluation' above and 'Lymph node spread' above.)

•For patients with suspicion for hilar or mediastinal (N2) adenopathy on chest CT, we suggest confirming this by endobronchial ultrasound biopsy/mediastinoscopy prior to attempted resection and not proceeding with resection if N2 disease is confirmed for histologies other than renal cell cancer. For patients with renal cell cancer and involved N2 lymph nodes, the decision to pursue complete metastasectomy must be individualized. (See 'Lymph node spread' above.)

•Contralateral mediastinal nodal involvement represents a contraindication to potentially curative metastasectomy for any histology. We also consider ipsilateral mediastinal involvement to represent a contraindication to pulmonary metastasectomy for tumors other than renal cell cancer since outcomes are generally poorer compared with involvement of just the intrapulmonary or hilar nodes.

•Prior to attempted resection of apparently isolated pulmonary metastases, brain imaging should be performed for patients who have tumors that frequently metastasize to the brain (eg, breast cancer, melanoma). (See 'Brain imaging' above.)

●Surgical approach

•For patients meeting criteria for pulmonary metastasectomy, we suggest using video-assisted thoracoscopic surgery (VATS) rather than an open thoracotomy for patients with one or a limited number of small metastases in the peripheral one-third of one lung (Grade 2C). (See 'VATS procedure' above and 'Open thoracotomy' above.)

•For patients planned for open thoracotomy, we suggest delaying surgery for at least six to eight weeks, rather than a shorter interval, to assess lesion stability and to avoid unnecessary surgery in those with rapid progression (Grade 2C). The need for delay in patients planned for a VATS procedure is less clear. We frequently choose not to delay a VATS metastasectomy for an isolated peripheral nodule in a favorable location. (See 'Open thoracotomy' above and 'Timing of surgery' above.)

•Prior to planned resection, we suggest evaluation of the mediastinal lymph nodes identified on chest CT that are >1 cm or PET-positive nodes of any size as would be done for a patient with a primary lung cancer (Grade 2C). This can be accomplished either with endobronchial ultrasound-guided biopsy or mediastinoscopy prior to surgery. We also agree with guidelines from the STS that recommend regional nodal sampling for evaluation of nodal metastases prior to attempted pulmonary metastasectomy for clinically node-negative carcinomas associated with a higher frequency of nodal disease spread (ie, renal cell, breast, colorectal carcinoma, and melanoma). (See 'Operative staging' above.)

●Postoperative management

•The role of postoperative adjuvant chemotherapy for specific histologic tumor types is discussed elsewhere. (See "Surgical resection of pulmonary metastases: Outcomes by histology".)

•Following initial complete resection, we suggest regular radiographic surveillance with chest CT scans every four months. (See 'Follow-up surveillance' above.)

•Radiographic evidence of possible recurrence mandates a full evaluation of disease extent and cardiorespiratory reserve. If the indications for pulmonary metastasectomy are still met, a repeat surgical procedure can be offered as an alternative to systemic therapy or other locally ablative options. A chest CT scan should be repeated six to eight weeks after the initial observation of suspected recurrence and prior to repeat surgery, as this may reveal additional deposits of recurrence that may preclude surgery. (See 'Repeat metastasectomy' above and 'Criteria for considering a resection' above.)

ACKNOWLEDGMENTS — The UpToDate editorial staff acknowledges Carlos Mery, MD, MPH, Carlos E Bravo Iniguez, MD, Aaron Dezube, MD, and Bryan M Burt, MD, who contributed to earlier versions of this topic review.