As long as surgery remains the best curative treatment for lung cancer, patients will continue to require pneumonectomy to treat lung cancer and for other occasional problems.1 Arguably, no other surgery carries as high a risk for perioperative mortality as pneumonectomy. Operative mortality from pneumonectomy has been reported to be between 5% and 20%.2–8 In a meta-analysis of 27 studies, 90-day mortality for right pneumonectomy was 20% and left pneumonectomy was 9%, for an overall mortality of 11%.9 For this reason, appropriate selection, operative technique, and postoperative management of patients who potentially may undergo pneumonectomy is crucial. We say potentially because patients scheduled for pneumonectomy ultimately may undergo a sleeve resection or exploration without resection depending on the findings at surgery.
Surprisingly, low preoperative lung function has not been demonstrated consistently to increase the perioperative risk of pneumonectomy, although some authors have found preoperative lung function to be an important factor.10–12 This finding may result from diligent efforts to identify and eliminate patients with poor pulmonary function from the surgical pool. Some have found poor preoperative lung function to increase the preoperative risk.13–15 Other factors have included increased age,7,8,11,14,16,17 right-sided procedures,8,9,12,14–19 preoperative chemo/radiation,12,14,20 large intraoperative fluid volumes,12,20,21 perioperative cardiac dysrhythmias,16 and immediate preoperative smoking history.12,22
This chapter includes discussion of the preoperative evaluation and management of pneumonectomy patients, the decision to perform pneumonectomy rather than sleeve resection, the technical aspects of the operation, and the postoperative management, all with the goal of decreasing perioperative mortality.
Preoperative Evaluation and Management
All patients with lung cancer, especially those who may undergo pneumonectomy, should first undergo preoperative staging (Table 72-1). At this time, a complete history and a physical examination that focuses on the identification of lymph nodes and liver masses, followed by a chest CT scan and a PET scan, are appropriate. This evaluation will rule out distant metastases other than brain metastases. Patients without symptoms of headache and with PET-negative mediastinum are unlikely to have brain metastases, so that brain CT scans or MRIs are not obligatory. However, the risk of the procedure supports appropriate evaluation to rule out brain metastases if the surgeon desires.
Candidates for pneumonectomy typically have large or hilar masses and thus a high likelihood of mediastinal metastases. Although PET scans are quite sensitive for detecting mediastinal metastases, they remain less accurate than mediastinoscopy23 (see Chapter 70). For this reason, tissue biopsy (either with endobronchial ultrasound, preoperative mediastinoscopy, or thoracoscopy) is indicated for patients who may undergo pneumonectomy, even if they have a PET-negative mediastinum. PET scans are useful for ruling out distant metastases, but should not be the only study used to evaluate the mediastinum.
We have discussed the importance of surgical staging of the mediastinum prior to pneumonectomy. Two other surgical evaluations (bronchoscopic and thoracoscopic) should be considered.
Bronchoscopic evaluation of the airway will ensure that a patient with a hilar mass on the right does not also have an endobronchial lesion on the left. The patient could die from the surgery to attempt to cure his right-sided cancer, and even if the surgery is successful, the pneumonectomy would have no impact on the left-sided cancer.
Thoracoscopic examination and staging of the pleural space benefits patients undergoing pneumonectomy in several ways24:
Thoracoscopy helps to rule out pleural metastases that usually cannot be identified except at exploration.
Thoracoscopic nodal sampling or lymphadenectomy can evaluate suspicious nodes in patients who have not had tissue nodal sampling.
Thoracoscopy can be used to determine the incision used for resection (see also Chapters 2 and 70).
An extensive but complicated literature describes the various modalities used to evaluate preoperative lung function as a means to predict postoperative complications. A preoperative forced expiratory volume in 1 second (FEV1) of more than 2 L has been relatively arbitrarily chosen as a threshold for resection without further pulmonary evaluation and has been confirmed by experience.
The obstructive nature of cancers that may lead to pneumonectomy assures that many patients will have less than 50% perfusion to the affected lung. Quantitative perfusion scanning has been used successfully for almost 30 years to determine the relative contribution of each lung to the patient’s lung function.6 Patients with a predicted postoperative FEV1 (better than preoperative FEV1) of more than 800 mL are considered to have adequate pulmonary reserve to undergo resection. This technique has been found to be accurate in long-term follow-up (up to 5 years),25 especially with respect to predicting the postoperative FEV126 (Table 72-2). Kim et al.13 have reported that perfusion less than 35% to the resected lung diminishes the operative risk. These parameters are also currently used in most cooperative group trials including lung resection.
Typical cardiac ischemia evaluations (i.e., stress testing) should be done in all patients with any history of cardiac ischemia (angina, other chest pain, or history of coronary artery bypass) and should be considered in patients with significant peripheral vascular disease.
Left pneumonectomy generally is safer than right pneumonectomy, at least in terms of overall perioperative and postoperative mortality. On the other hand, risks associated with long bronchial stump require careful technical attention to details of resection. However, the procedure may be technically more difficult, especially in patients with large hilar masses.
Left hilar cancers are truly in “tiger country,” approaching or involving the aortic arch, recurrent laryngeal nerve, phrenic nerve, pericardium, and the main pulmonary artery before it bifurcates into right and left main pulmonary arteries. (The left main pulmonary artery is very short.) The primary procedures are described separately, since the anatomy and the dissections are different on the left and right sides.
The mediastinal pleura is divided circumferentially around the hilum using either scissors or an energy probe, including the inferior pulmonary ligament. Harvesting the lymph nodes in the aortopulmonary window gives better exposure to the proximal pulmonary artery. This must be done carefully to avoid injuring the recurrent laryngeal nerve, which branches off the vagus nerve on the inferior surface of the aortic arch and then dives into the mediastinum immediately inferior and adherent to the arch. Unless it is involved with tumor, the phrenic nerve should also be preserved until a decision is made about sleeve resection. It may abut the anterior aspect of the nodes in the AP window, and so is also at risk.
Harvesting the subcarinal lymph nodes gives better access to the bronchus and allows the surgeon to better palpate the bronchus. Cautery should be used judiciously or not at all in this space, since the blood vessels to the bronchus arise here.
The vagus superior to the aorta or the recurrent nerve can be harvested if either is involved with cancer. However, resection of these nerves will lead to vocal cord paralysis.
The lung is examined carefully to determine whether sleeve resection is possible. Maneuvers that help include:
Dividing the pleura circumferentially;
Taking down the inferior pulmonary ligament;
Complete parenchymal separation of the upper and lower lobes; and
Completing the mediastinal lymphadenectomy.