Pulmonary metastases are defined as secondary malignant tumors of the lung parenchyma or pleura. Originating from an entirely separate organ, metastatic disease to the lung is the second-most common pulmonary malignancy. Autopsy studies have demonstrated that 20 to 54 percent of patients who die with a diagnosis of cancer also have pulmonary metastases. A subset of these patients have metastases limited only to the lung. In their study of 162 sarcoma patients, Huth and Eilber reported that of the patients with recurrent disease, 88 percent had pulmonary metastases, with the lung being the sole site of disease in 72 percent.¹ Similarly, Billingsley and associates examined 230 patients with recurrent soft tissue sarcomas treated at Memorial Sloan–Kettering Cancer Center and found that 73 percent of these patients had recurrences that appeared initially only in the lungs.² With a significant proportion of patients with metastatic malignancy confined to the lung, complete resection of the intrathoracic disease burden would render these patients disease-free.

Although the majority of operations for pulmonary metastasectomy can be performed via an open approach, thoracoscopic pulmonary metastasectomy has been demonstrated to be equally efficacious with similar survival rates and has the benefit of fewer adhesions, less pain, and shorter length of stay.

Weinlechner and colleagues published the first report of pulmonary metastasectomy in 1882.³ In 1927, Tudor Edwards performed a sublobar resection for recurrent sarcoma following an amputation 6 years earlier. Barney and Churchill described a lobectomy after nephrectomy for synchronous metastatic kidney cancer in 1933; the patient subsequently survived for 23 years.⁴ Blalock described the first pneumonectomy for metastatic large bowel carcinoma at a “recent advances” lecture in 1944.⁵

In 1947, Alexander and Haight published a case series of pulmonary metastasectomies, proposing the first set of selection criteria for surgery.⁶ Since then, pulmonary metastasectomy has become a widely accepted treatment modality for patients with metastatic disease. Recent series have described 5-year survival rates of 40 percent in osteogenic sarcoma,⁷ 38 percent in soft tissue sarcoma,⁸ 40 percent in colorectal carcinoma,⁹ and 50 percent in patients with breast carcinoma.¹⁰

Stephen Paget first described his “seed and soil” hypothesis in 1889. He observed that metastatic disease fell in a nonrandom pattern and postulated an affinity of certain primary cancer types (the “seed”) toward factors within the metastatic target organ (the “soil”). In 1928, James Ewing challenged this hypothesis, proposing that metastasis was directly related to the anatomy of the vascular system instead. Regardless of which of these two hypotheses holds true, new improvements in cellular, molecular, and imaging techniques have elucidated that the development of metastatic lesions is the result of a complex, multistep process known as the metastatic cascade.¹¹ In order for a metastatic focus of cancer cells to form, cells from a growing tumor need to escape the local environment (intravasation, resistance to anoikis), travel while avoiding host defenses (evasion of the immune system), leave the bloodstream (extravasation), and begin the process of implantation and growth (dormancy and proliferation).

Recent advances in identifying the gene expression profiles of varying tissues have led to investigations in identifying a signature of genes characteristic of metastatic tumors. Gene expression profiling studies using cDNA microarrays comparing the expression profiles of primary tumors with the expression profiles of metastatic tumors in patients with breast cancer have identified a 70-gene signature that successfully differentiated tumors with poor versus good prognosis.¹² Ramaswamy and coworkers identified a metastatic signature based on gene expression profiling that could predict the presence of metastases across a variety of adenocarcinomas.¹³ Lastly, although similar studies in humans are yet to be performed, Minn and colleagues were able to identify an expression signature characteristic of pulmonary metastases in a murine model.¹⁴ Undoubtedly, understanding the metastatic cascade and elucidating the mechanisms of each step will provide insight into potential treatments against metastatic disease.

Secondary tumors of the lung are most often an occult finding during surveillance imaging after primary tumor resection. Patients are typically asymptomatic. Rarely, patients with pulmonary metastases will present with symptoms of dyspnea, cough, hemoptysis, or pain. Pleural involvement may result in pneumothorax or pain. Thorough history taking may elicit a history of unintentional weight loss, general malaise, or decreased energy. Physical examination is likely to be negative for significant findings.

Although plain chest radiography was previously the mainstay of screening for pulmonary metastases, computed tomography (CT) has emerged as the most commonly used screening imaging modality and is the current standard. On CT imaging, pulmonary metastases appear as single or multiple pulmonary nodules within the lung parenchyma. In general, these lesions are rounded and well circumscribed. Hilar or mediastinal lymphadenopathy may also indicate lymphatic involvement. In some cases, the pulmonary nodule(s) may contain tissue elements consistent with the primary site (e.g., calcification in metastatic osteosarcoma or fatty elements in liposarcoma).

High-resolution CT (HRCT) can detect pulmonary nodules as small as 3 mm. In a study comparing conventional CT with HRCT, Margaritora and colleagues reported a sensitivity of 100 percent for both types of scans for nodules 10 mm in size.¹⁵ The sensitivity of nodules smaller than 6 mm dropped to 62 and 48 percent with conventional CT and HRCT, respectively. Overall sensitivity for pulmonary metastases was 82.1 percent. In examining the sensitivity of thin-slice CT, Pfannschmidt and coworkers found the overall sensitivity of 3- and 5-mm slices to be 88.8 and 83.7 percent, respectively.¹⁶ The sensitivity decreased to 74.7 (3-mm slices) and 64.0 percent (5-mm slices) for nodules smaller than 5 mm.

The use of positron emission tomography (PET) scan in the detection of pulmonary metastasis has also been studied. Fortes and colleagues found an overall sensitivity of 67.5 percent in the detection of pulmonary metastases with PET imaging.¹⁷ For lesions 10 mm or greater in size, the sensitivity of PET was 87.8 percent; however, for lesions smaller than 10 mm, sensitivity was only 29.6 percent. It is also notable that the sensitivity varied with histology. For metastases with squamous cell carcinoma histology, sensitivity of the exam was 93 percent. For sarcoma, the sensitivity of PET was 44 percent. Although the sensitivity of PET imaging is variable and this modality should not be used solely for the screening of pulmonary metastases, it does serve as an adjunct to CT imaging and is particularly useful in evaluating for the presence of extrathoracic disease.

Chemotherapy

Systemic chemotherapy should be used for patients with widely metastatic disease or as an adjunct to surgical therapy in patients with metastatic germ cell tumor or breast cancer. Patients with a solitary pulmonary metastasis or oligopulmonary metastases should be considered for surgical resection, as 5-year survival following resection can be as high as 40 percent.¹⁸ Patients who cannot tolerate surgery or refuse surgery for metastases confined to the lung should be considered for other forms of locoregional therapy such as radiofrequency ablation (RFA), stereotactic body radiation therapy (SBRT), or isolated lung perfusion (ILuP).

Radiofrequency Ablation

RFA uses an alternating current to create frictional heating of tissue, resulting in tissue necrosis. A radio wave—emitting electrode is inserted into the tumor, which is then heated to 90°C to 100°C to produce coagulation necrosis. Originally used in the treatment of hepatic tumors, the use of RFA was soon extended to the treatment of pulmonary malignancies. Candidates for RFA should have disease confined to the lung(s) or otherwise controlled. In addition, lesions amenable to RFA should be small in diameter (less than 3 cm in diameter) and should not abut mediastinal structures, which may act as a heat sink.¹⁹

In general, RFA is associated with low morbidity and mortality. Common complications associated with RFA are usually mild and self-limited. Patients should be counseled regarding the possibility of pain during the procedure, pneumothorax (possibly requiring thoracostomy tube management), fever, hemoptysis, cough, pleural effusion, hemothorax, and abscess formation in the ablation bed.

Lesion size is an important consideration of RFA for treatment of metastatic disease. Several authors have demonstrated statistically significant differences in response rates as well as overall survival for lesions greater than 3 cm versus those smaller than 3 cm.^20,21 In a study of RFA for renal cell carcinoma metastases, the 5-year survival of patients with fewer than six nodules that were less than 6 cm in diameter was 100 percent. Recurrence-free survival was 92 percent at 1 year and 23 percent at 5 years. Another study of 22 patients with metastases of varying histology demonstrated a 2-year survival of 68 percent. Lesion size was found to be a significant prognostic factor in both recurrence and survival.¹⁹

One important aspect to remember regarding RFA is that since there is no pathologic analysis, the treatment margin is unknown. A study of RFA of 18 pulmonary lesions that were subsequently resected for pathologic analysis demonstrated only 39 percent of lesions were completely ablated and only 50 percent of patients had more than 90 percent tumor ablation.²² More importantly, 11 percent of patients had less than 90 percent tumor ablation. Although the authors contend that early pathologic analysis may have underestimated the effectiveness of tumor ablation, these results are inferior to those achieved with surgical resection. Therefore, RFA should be reserved for treating pulmonary metastasis in patients who are medically inoperable or who refuse surgery.

Stereotactic Body Radiation Therapy

SBRT is a technique that uses precisely targeted radiation. This technique facilitates the delivery of higher doses of radiation to the target anatomy, resulting in better tumor control, while minimizing harmful effects to adjacent normal tissue. Patients being considered for SBRT should meet the same criteria as those being considered for RFA. Although data regarding SBRT for pulmonary metastases are limited, a multi-institutional Phase I/II trial of SBRT for pulmonary metastases demonstrated local control of 100 and 96 percent at 1 year and 2 years, respectively, and a median survival of 19 months.²³

Compared with conventional radiotherapy, SBRT has several unique features. First, the doses of radiation administered are high and localized. To reduce treatment toxicity, there are tight dose distributions around the tumor, with rapid dose falloff in the surrounding lung parenchyma. During planning and treatment, the radiation oncologist can account for tumor motion. Finally, there is a reproducible patient immobilization to minimize movement during treatment sessions. Complications associated with SBRT affect up to 5 percent of patients, and most commonly are pneumonitis (grade 2 and higher) and rib fractures.

A significant limitation to SBRT is that the treated lesions should be small in diameter. In addition, the treatment of multiple lesions presents as a challenge while attempting to minimize toxicity to adjacent, normal lung parenchyma. As such, although SBRT should not be the primary modality used in treating secondary pulmonary malignancies, it can serve as a good alternative in treating poor surgical candidates or as an adjunct to surgical therapy in patients with multiple lesions.

Surgical Metastasectomy

Surgical resection is the mainstay of therapy for patients with secondary pulmonary malignancies in the absence of extrapulmonary disease and controlled primary disease. Alexander and Haight proposed the first set of selection criteria, listed in Table 10-1, for patients eligible for pulmonary metastasectomy.⁶ First and foremost, the site of primary disease must be either controlled or controllable. Ideally, the patients should also have an absence of extrapulmonary disease. For patients with an additional site of metastatic disease outside the lung, the extrapulmonary disease should be controlled or controllable. Finally, the patient should have sufficient pulmonary reserve to tolerate the loss of lung parenchyma required to achieve a complete resection. With the appropriate selection of patients, 5-year survival rates of 20 to 40 percent have been reported.¹⁸

Patients who are candidates for resection may be treated with either open approach or thoracoscopic approach. Proponents of open resection state that open surgery is mandatory for sufficient manual palpation of the lung parenchyma. However, supporters of thoracoscopic surgery assert that the sensitivity of HRCT obviates the need for manual palpation and that the benefits of a thoracoscopic approach outweigh those of an open approach. Compared with thoracotomy, benefits of thoracoscopy include decreased postoperative pain, decreased chest tube duration, decreased hospital length of stay, faster recovery, and decreased inflammatory response facilitating re-do surgery.

In an attempt to provide guidelines for surgical approach in pulmonary metastasectomy, the European Society of Thoracic Surgeons working group carefully evaluated the considerations involved in decision-making.²⁴ They narrowed the issue of surgical approach to four questions: (1) What is the evidence for the need for palpation of the lung in the modern era of imaging? (2) Is there evidence of a difference in outcome for an open versus a closed approach? (3) Is there evidence for a difference in outcome for an initial policy of bilateral versus unilateral exploration? (4) In patients with known bilateral disease, is there a difference in outcome with simultaneous versus a staged approach? Each of these topics is an important consideration in the surgical approach to pulmonary metastases.

As previously mentioned, HRCT has an overall sensitivity of 88.8 and 83.7 percent for 3 and 5 mm slices, respectively.¹⁶ Although the sensitivity decreased to 74.7 (3-mm slices) and 64.0 percent (5-mm slices) for nodules less than 5 mm in size, it is difficult to argue that manual palpation has a sensitivity superior to HRCT scans currently available. However, in early studies investigating the need for manual palpation of metastatic disease, 10 of 15 patients had additional nodules that were not identified on CT scans or during their video-assisted thoracic surgery (VATS) procedures.²⁵ More recent studies have similarly reported the identification of additional nodules during thoracotomy that were not recognized either on CT imaging or during thoracoscopy.^26–28

Although an open approach may identify more nodules than during thoracoscopy, this finding does not translate to improved survival. In a review of seven studies by Molnar, six demonstrated no difference in survival between patients undergoing metastasectomy by VATS versus open surgery.²⁴ Moreover, patients undergoing thoracotomy experience significantly more complications than those who have similar parenchymal resection by thoracoscopy.²⁸ In a recent cohort study comparing patients treated with metastasectomy via VATS versus thoracotomy, Carballo and associates demonstrated a 5-year overall survival of 58.8 percent for thoracotomy and 69.6 percent for VATS (p = 0.03).²⁹ There were no statistically significant differences between the two groups in terms of recurrence-free survival.

Finally, there are no data demonstrating a difference in outcome between an initial policy of bilateral versus unilateral exploration and simultaneous versus staged approach in patients with known bilateral disease. Results from the European Society of Thoracic Surgeons working group survey suggested that an initial approach via median sternotomy is reasonable for patients with known bilateral disease.²⁴ In cases not suitable for median sternotomy, such as posterior lesions or patients with previous pulmonary resection, staged thoracotomy with a 3- to 6-week interval was recommended. Zheng and colleagues recommended a similar approach to pulmonary metastasectomy except that these authors were more supportive of a thoracoscopic approach.¹⁹

In our experience, complete thoracoscopic resection of pulmonary metastases is possible in 75 percent of cases. We have resected as many as six ipsilateral nodules in a single setting, and have routinely had success in resecting nodules less than 5 mm in diameter. Figure 10-1 depicts an algorithm for managing pulmonary metastasectomy.

Ultimately, the main tenet of surgical resection of pulmonary metastases is complete resection. Peripheral nodules should be treated with wedge resection; more central nodules may require a segmentectomy, lobectomy, or, rarely, pneumonectomy. With regard to open versus thoracoscopy, “there is such evident variation that depends on the particular needs of the patient and the preference of the surgeon that it might well be a situation in which it is a decision for individual craftsmen to do what works best in their hands.”²⁴