The type of surgical approach used in pulmonary metastases is highly variable. The choice of approach varies between thoracic surgery departments and between countries. This fact was noticed in the survey of the European Society of Thoracic Surgeons undertaken in 2006 [1]. The survey showed that palpation of the lung was regarded as mandatory by 65% of responders, but the use of videothoracoscopic surgery (without complete palpation) was acceptable to 60%.

Debate in the surgical literature exists over whether pulmonary metastases should be removed by a thoracoscopic approach or by thoracotomy, due to the high rate of non-imaged metastases. Cerfolio et al. found that 52 (34%) of 152 patients had non-imaged pulmonary nodules detected by palpation; almost half of these nodules were malignant [44]. Eckardt prospectively evaluated 37 patients who underwent video-assisted thoracoscopic surgery (VATS), followed by thoracotomy, in whom 29 additional nodules of a total of 84 were resected (35% non-imaged nodules). Of non-imaged resected nodules, eight (28%) were malignant [45].

Recently, the European Metastasectomy Working Group made a systematic review on this issue [46]. They reviewed seven studies providing outcome data comparing VATS with thoracotomy, with no survival difference between groups in six studies; in one of them, they found lower recurrence-free survival at 5 years in patients having open surgery (21%) rather than videothoracoscopic surgery (34%). These authors admitted to disagreement within the working group, but still concluded lung palpation was necessary [46].

With regard to the choice of approach for bilateral disease based on the ESTS survey [1], two out of three surgeons preferred bilateral staged thoracotomy, followed by sternotomy, bilateral sequential thoracotomy (one stage), bilateral staged thoracoscopy, bilateral thoracoscopy (one stage), and finally clamshell incision. In the 1980s and 1990s, several authors published their results in the use of sternotomy for pulmonary metastasectomy, reporting higher accuracy for detecting unsuspected metastases and lower morbidity [47–50]. The utility of a bilateral approach to discover unsuspected disease may progressively diminish in importance as imaging improves [46].

In current practice, pulmonary metastasectomy may be performed more than once; therefore, conservative resections are recommended. Preservation of as much functioning lung parenchyma as possible is an agreed principle, while removing a centimeter of the surrounding pulmonary tissue to ensure free resection margins [51]. For nodules located peripherally, stapled wedge resection is generally the preferred treatment (Fig. 21.1) but for large or central lesions, segmental resections, lobectomy, or occasionally, pneumonectomy may be required [51, 52]. In patients in whom the planned resection is deemed to be complete, extended resections beyond lung parenchyma, including the chest wall, the azygos vein, the diaphragm, the pulmonary vessels, and larger lung resections, such as sleeve lobectomies and pneumonectomies, are associated with 5- and 10-year survival rates of 42% and 36% respectively [53].

For those patients with a low cardiopulmonary reserve, cautery resection, or laser resection have been proposed as an alternative to save as much surrounding lung as possible.

Cautery resection or precision resection was described by Perelman [54]. This technique allows excising deep-seated lesions or nodules located on the broad surface by coring them out of the parenchyma using cautery (Fig. 21.2). This technique is useful for those cases with multiples metastases or in patients who are not considered good candidates for anatomic resection.

Laser resection uses the same principles of cautery resection with the following major advantages: it permits limited excision of deep-seated lesions sparing lung tissue as much as possible, there is minimal deformity or damage to the adjacent lung tissue, and for lesions located near a major bronchus or vessel a maximum margin of tissue around the lesion can be taken without injury to these adjacent structures [51]. In 2006, Rolle et al. [55], the group that has dedicated more efforts to the development of this technique, published their results on 328 patients with a 5-year survival rate of 41% following laser resection of a mean of eight metastases per patient (range 1–124). They recommended laser resection even in patients with more than 20 metastases, because a 5-year survival rate of 26% was observed [55].

With regard to the extent of resection, Migliore et al. reviewed seven reports addressing pneumonectomy for metastatic disease, five of whom reported completion pneumonectomy [56]. Operative mortality rates ranged between 0% and 11% (19% in one study with R1 resections) and 5-year survival rates ranged between 10% and 41%. Seven studies reviewed the issue of repeat metastasectomy with 5-year survival rates ranging from 19% to 53.8%. Thirty percent to 40% 5-year survival rates are reported after a third or fourth thoracotomy. Resectability was an independent prognostic factor for survival after each subsequent thoracotomy, but the chance of being resectable decreases after each thoracotomy, as does the chance for permanent control. After patients are determined to be unresectable, median survival was 8 months (19% 2-year survival rate) [56].

Postoperative complications are rare, occurring in between 0% and 22% of patients [4], and the most common are respiratory (pneumonia and acute respiratory syndrome), cardiovascular (arrhythmias), and those related to surgical technique (air leaks) [2]. Postoperative mortality rates range between 0% and 2.5% [4], and are lower for sublobar and lobar resections than for pneumonectomy. Reported causes of death are pulmonary embolism, pneumonia, respiratory failure, and cardiac failure [2, 4].

The role of lymph node assessment in pulmonary metastasectomy has not been well defined, and thus it is not a common practice in most thoracic surgery departments around the world. The results of the survey of the European Society of Thoracic Surgeons with regard to lymph node assessment at the time of pulmonary metastasectomy showed that 55.5% of the responding surgeons performed mediastinal lymph nodal sampling, while 13% performed a complete mediastinal lymphadenectomy. One surgeon out of three (32.2%) performed no lymph node biopsy whatsoever [1].

Regarding the prevalence of lymph node involvement, Garcia-Yuste et al. [57] analyzed six studies, finding a prevalence ranging from 14 to 32%, giving a weighted average of 22%. Tumors were of varied types but colorectal carcinoma was well represented.

Regarding the impact on survival, the meta-analysis published by Gonzalez et al. in 2013 [3], found that involvement of the hilar and/or mediastinal lymph nodes was associated with a clear increased risk of death, HR = 1.65 (95% CI 1.35–2.02). Recently, the Spanish Colorectal Metastasectomy Registry of the Spanish Society of Pneumology and Thoracic Surgery has examined the impact on survival depending on the pathological nodal status based on data from 522 patients collected prospectively from 2008 to 2010 [58]. For this purpose, a systematic nodal dissection or nodal sampling was required to certify pathological absence of lymph node metastases, or the pathological lymph node status was coded as uncertain. The 3- and 5-year disease-specific survival rates of patients with lymph node metastases and uncertain lymph node status were clearly independent, with a worse survival in comparison with the survival of patients with no lymph node metastases; presence of lymph node metastases was 50.5% (95% CI 29.6–71.4)/24.8% (95% CI 3.6–46) respectively; uncertain lymph node status, 69% (CI 64–74.4)/44% (CI 35–53) respectively; and absence of lymph node metastases: 73.5% (95% CI 65–82)/58.3% (95% CI 41.6–75.1) respectively. With these results, the Spanish group concludes that the presence of lymph node metastases remains an important prognostic factor, and hypothesizes that the missed lymph node metastases in patients with uncertain lymph node status can impair survival because of misclassification of risk [58]. Awaiting a future randomized trial to answer the precise impact of surgery in lung metastases from CRC, the current findings in the literature suggest that intraoperative nodal assessment should be performed in metastasectomies, at least to determine the individual postoperative prognosis.

Despite more refined and extensive resection techniques, enhanced selection of patients, and evolving multidisciplinary treatment options, only a small proportion of patients with isolated pulmonary metastases undergo resection. Systemic chemotherapy has become an important treatment modality for metastatic colorectal cancer. For advanced disease, combination chemotherapy consisting of 5-fluorouracil (5-FU), leucovorin, and oxaliplatin (FOLFOX), or 5-FU, leucovorin, and irinotecan (FOLFIRI) is most widely used, with median overall survival ranging between 14.7 and 20.0 months [59–61]. By addition of cetuximab in selected patients, median overall survival time increases to 23.5 months [61].

In the case of liver metastases that are considered for surgical resection, perioperative combination chemotherapy with the FOLFOX regimen is recommended [59]. However, the optimal treatment schedule has not been determined yet.

In patients undergoing pulmonary metastasectomy, the role of induction or adjuvant chemotherapy has not been established, and no convincing evidence is currently available supporting its use in daily practice. For this reason, every patient with lung metastases from colorectal cancer should be discussed within a multidisciplinary board including a thoracic oncologist and surgeon to determine optimal, individualized treatment.

As already shown in the large retrospective database of Pastorino et al., most patients who underwent macroscopic complete resection of pulmonary metastases from colorectal cancer or sarcoma will develop single or multiple recurrent disease inside the chest [13].

The dose of intravenously administered chemotherapy is limited by systemic toxicity. Regional drug delivery systems may enhance drug uptake in lung tissue and minimize systemic side-effects and toxicity [62, 63]. Several techniques are able to administer a high dose of loco-regional chemotherapy very efficiently, of which isolated lung perfusion, selective pulmonary artery perfusion, and chemoembolization are the most thoroughly investigated.