Preoperative Staging and Surgical Approaches
Preoperative Assessment
The pulmonology perspective
Although surgeons generally determine the resectability of a lung tumor, operability is determined by an assessment of the patient's physiology. Preoperative assessment focuses on the likelihood of complications and prediction of post-surgical lung function. When evaluating patients for lung resection, it is important to consider that a pneumonectomy often results in loss of one third of lung function and 20% reduction in exercise capacity. A lobectomy, by comparison, usually results in the loss of less than 10% of lung function and preservation of exercise capacity. Traditionally, most preoperative assessments use pulmonary function tests and some form of exercise testing. Pulmonary function testing unfortunately is of limited predictive value for the individual patient. A forced vital capacity (FVC) of less than 70% and a predicted postoperative forced expiratory volume in 1 second (FEV1) of less than 800 mL are often used as parameters for predicting significant postoperative problems. Interestingly, baseline hypoxia or hypercapnia are not proven to prohibit surgical intervention. Exercise testing, such as a formal cardiopulmonary exercise test or stair climbing exercise, has become a popular tool. If patients can climb 8 flights of stairs at their own pace without holding on, they are unlikely to have severe postoperative complications. Fewer flights are predictive of a rising incidence of complications. For example, the ability to only climb 5 flights roughly equals an exercise capacity associated with an oxygen consumption per unit time of 20 mL/kg/min and is considered to predict postoperative complications.
Rather than adhering to formulas, it is best to approach each patient individually. If there is little impairment in pulmonary function and exercise tolerance appears reasonable, no further workup is necessary. Patients who are impaired should be assessed by a pulmonologist for risk stratification. Additionally, this referral should be used to optimize medical treatment before surgery and to initiate an exercise and rehabilitation program. In clinical practice, it is not unusual that with an aggressive and supervised effort like this, patients who were deemed high risk or to have inoperable disease can safely undergo curative surgery for lung cancer.
Staging and the Role of Mediastinoscopy
The surgical perspective
For early-stage disease, staging assessment should include a diagnostic chest computed tomographic (CT) scan, a positron emission tomographic (PET) scan to evaluate the mediastinal lymph nodes and exclude extrathoracic or bone metastases, and liberal use of mediastinoscopy in most patients. Lymph node metastasis portends a worse prognosis, and surgical resectability for cure is greatly influenced by mediastinal lymph node involvement. Evaluation of mediastinal lymph node metastases by CT alone has been extensively reviewed in the literature and has historically yielded an overall sensitivity of 59% (range, 20%-81%) and an overall specificity of 78% (range, 44%-100%)14 PET evaluation of the mediastinum for stage I and II NSCLC has yielded a sensitivity of 85% and specificity of 88%15 Its negative predictive value for the mediastinum is 87%, whereas the positive predictive value is 56%. Meta-analysis of 17 relevant studies that explored the diagnostic accuracy of PET compared with CT imaging in detecting mediastinal lymph node metastases demonstrated that PET was more accurate.16 Accurate detection of hilar nodal involvement in lung cancer is much more difficult, and neither PET nor CT accurately identifies nodal disease in the hilum.
The increasing use and availability of PET CT fusion imaging are increasing the proportion of patients shown to have additional lesions that could contraindicate surgical treatment but that require tissue confirmation to exclude a false-positive examination result. In a frequently cited prospective study from Switzerland, Lardinois and colleagues17 studied this technology in 50 patients with NSCLC. Integrated PET CT provided additional information in 41% of patients and was significantly more accurate in precise staging compared with CT alone, PET alone, or visual correlation of PET and CT.
The gold standard for staging the mediastinum is cervical mediastinoscopy, which is the modality on which all comparisons of accuracy are currently made. The efficacy of mediastinoscopy has been well established. In patients with known or suspected lung cancer, the routine use of mediastinoscopy can change the plan of care in up to 25% of patients. Large studies confirm low morbidity and mortality as shown in Table 1.18-20
Despite its established efficacy in staging, many lung cancer treatment centers use mediastinoscopy only selectively. A national survey of 729 hospitals (31% teaching or university hospitals, 38% community cancer centers, 46% comprehensive community cancer centers) reported that only 27% of 11,668 patients whose initial management included surgical therapy for lung cancer were evaluated preoperatively with mediastinoscopy.21 The underuse of this modality is striking. Additionally troublesome is that only 42% had lymph nodes sampled at any mediastinal level during surgery.
Mediastinoscopy should be performed in any patient with enlarged or PET-avid mediastinal lymph nodes, those with central tumors, and those with peripheral tumors 3 cm or larger. Peripheral T1 lesions 1 to 2 cm or smaller with PET-negative mediastinal lymph nodes can be regarded as the one potential exception to the routine use of mediastinoscopy. Tumors larger than 2 cm are twice as likely to have mediastinal lymph node metastases compared with tumors smaller than 2 cm.
Mediastinoscopy has some limitations, including the need for general anesthesia and the inability to reach all nodal stations (Figure 3). These limitations have fostered the development of additional modalities to enhance lung cancer staging, including endobronchial ultrasonography (EBUS) and esophageal ultrasound-guided fine needle aspiration (EUS FNA). These technologies are not necessarily mutually exclusive and can be considered complementary in the comprehensive evaluation of mediastinal lymph nodes.
EBUS-Guided TBNA and EUS FNA
With few exceptions, most lymph nodes are grouped alongside the airways or the esophagus. It is therefore intuitive that far more lymph node stations should be accessible using endoscopic routes. Additionally, endoscopy does not require the use of an operating room or general anesthesia, and it is exceedingly safe and relatively inexpensive. Endoscopic lymph node sampling can also be repeatedly performed safely, unlike mediastinoscopy, which is difficult to perform repeatedly.
Transbronchial needle aspiration (TBNA) sampling, which uses a flexible needle to obtain a sample during a regular bronchoscopic procedure, was introduced in the 1980s. Although TBNA has an exceptional safety record, it has never been widely used because many practitioners believe the yields are suboptimal. This shortcoming is most likely because the lymph nodes are not visible with conventional bronchoscopy, and the needle biopsy has to be performed in an area that has been predetermined by reviewing the patient's CT scan.
A tremendous step in the right direction was undertaken with the introduction of miniaturized ultrasound probes for bronchoscopy and EBUS. Ultrasound performed during the procedure can visualize parabronchial structures and guide TBNA.22 A further improvement has been the introduction of dedicated ultrasound bronchoscopes, which can allow biopsies to be performed on lymph nodes as small as 6 mm. Reported yields for EBUS-guided TBNA are generally in the range of 75% to 85%, a vast improvement over the early days of "blind TBNA" but still largely dependent on the skill of the endoscopist.4 The advantage of EBUS-guided TBNA over mediastinoscopy is clear: the procedure is quickly performed with the patient under moderate sedation, is safe and inexpensive, and can be repeated if necessary. Additionally, it affords access to lymph nodes not easily harvested through mediastinoscopy, such as hilar stations and posterior carinal abnormalities, while providing easy access to nodes conventionally biopsied surgically.
Another development of note is EUS FNA. In many ways this technique is similar to bronchoscopic lymph node sampling but through the esophagus. Fine needle aspiration through this route is also a routine procedure similar to conventional esophagogastroduodenoscopy and performed on an outpatient basis. The medical literature supports similar yields to EBUS TBNA in experienced hands. EUS FNA allows for the sampling of station 8 nodes, celiac nodes, and adrenals, stations not accessible by bronchoscopy or medastinoscopy.23-24 EBUS TBNA in contrast reaches hilar nodes and anterior carinal nodes, stations not in reach for EUS FNA. These results have created significant interest, because it seems feasible that endoscopic staging with combined EUS FNA and EBUS TBNA may be an optimal way to complete staging in a patient with complementary tools in one setting with one sedation. In many institutions, efforts are under way to create endoscopy suites that support the combined use of both technologies.
Dr. Lynch: What kind of staging do you require?
Dr. Dupuy: The PET scan is used in almost everybody. There is a lot of literature supporting the role of PET because it can upstage and might change the operative approach, particularly in an older person, where you may elect to be more aggressive with treatment.
Dr. Lanuti: Mediastinoscopy is still the gold standard because of its low morbidity and low mortality. In staging patients surgically, we tend to get a PET scan on almost everybody, and fusion PET CT is something that's becoming much more prevalent, even in the community.
Dr. Dupuy: Physicians want to know if PET is appropriate when you have a spiculated mass. Even if the lesion hasn't been biopsied for staging, a PET scan can help determine if the lesion is cancerous. At least the patients will be staged appropriately with the least invasive approach.
Dr. Ernst: CT is obviously just part and parcel of this approach as well. Once a chest x-ray or a bronchoscopy identifies the mass, a PET or a PET CT is routine. There are some nice new technologies, such as confocal endoscopy and confocal imaging in the GI tract, which is basically microscopic imaging. With this technique, you get cellular resolution. So, rather than a full biopsy wedge, you have an optical biopsy. This raises the issue of who is going to read these results in the future and do endoscopy units need to be directly connected to the pathology department.
Surgical Approaches
Currently, all patients with clinical stage I (T1 N0, T2 N0) and most patients with stage II (T1 N1, T2 N1, T3 N0) NSCLC should undergo surgical resection as definitive primary therapy. Although early-stage lung cancer (stage I-II) is often resectable, the 5-year survival rates in North America remain disappointing, at 57% to 80% for stage I disease and 39% to 55% for stage II disease.25,26
Surgical techniques for the treatment of early-stage lung cancer include wedge resection, segmentectomy, lobectomy, pneumonectomy, and parenchymal-sparing sleeve resection. The standard resection for early-stage NSCLC in the past 50 years has been defined by anatomic lobectomy, which permits removal of the primary tumor along with the corresponding lymphatic drainage. Lesser resections, such as wedge excision and anatomic segmentectomy, have been regarded as parenchymal-sparing modalities for patients with compromised pulmonary function. In 1995, the Lung Cancer Study Group27 published a randomized trial of lobectomy compared with limited resection for the treatment of stage IA (T1 N0) NSCLC. The study revealed a 3-fold increase in the locoregional recurrence rate (6.5% vs 16.8%) among patients undergoing limited resection. There was no significant difference in 5-year survival rates (65% vs 44%) for those undergoing lobectomy compared with limited resection (Table 2). The authors concluded that lobectomy was the procedure of choice for patients with stage I NSCLC. This study established the foundation for lobectomy as the standard of care; however, it has been criticized by many investigators for the lack of routine staging chest CT and the inclusion of a large number (33%) of wedge resections over and above segmentectomy.
Since 1995, technical advances in CT have permitted the identification and characterization of both ground glass opacities and subcentimeter pulmonary neoplasms. This phenomenon has prompted a number of investigators to revisit the notion of limited resection for early-stage NSCLC. Keenan et al28 reported no statistically significant difference in overall (67% vs 62%;P= .80) and disease-free survival (82% vs 74%; P = .70) in patients with stage IA NSCLC treated by lobectomy or segmentectomy. As shown in Table 2, a prospective Japanese study by Okada and colleagues29 reported a 10-year experience comparing sublobar resection (segmentectomy or wedge) with lobectomy for NSCLCs 2 cm or smaller. Multivariate analysis confirmed that the recurrence rate and 5-year survival rate for sublobar resection were not inferior to lobar resection. In contrast, a large North American retrospective review30 found a higher local recurrence rate with limited resection compared with lobar resection, but no significant difference in disease-free survival at 5 years. These studies emphasize the need for a modern prospective randomized trial comparing limited resection with lobectomy for select patients who are surgically staged as having small peripheral stage IA NSCLC. This will be addressed in a cooperative phase 3 randomized trial sponsored by the Cancer and Leukemia Group B (CALGB 140503) specifically evaluating lobectomy vs sublobar resection for peripheral NSCLCs 2 cm or smaller.
Role of Video-Assisted Thoracoscopy
Thoracoscopy is an advancing modality for performing anatomic lung resections and confirming pathologic stage. Video-assisted thoracoscopy (VATS) can prevent the use of thoracotomy and theoretically reduce hospital stay and associated postoperative morbidity. It is frequently used at pulmonary resection to rule out pleural metastases, assess abnormal aortopulmonary lymph nodes, and more accurately assess T stage, particularly when the tumor abuts chest wall, vascular, or mediastinal structures (Figure 3).
A few large series now exist documenting that VATS lobectomy is associated with low perioperative morbidity and mortality.31-33 As experience matures, survival rates appear compatible with conventional thoracotomy and lobectomy. Advantages of VATS lobectomy include earlier return to daily activity, decreased hospital stay, decreased inflammatory response,34 less incidence of atrial fibrillation, and perhaps improved early postoperative pulmonary function. There are conflicting reports regarding improved postoperative pain after VATS lobectomy compared with limited thoracotomy,35 which need further study. To date, VATS lobectomy has not been widely accepted and is only performed in 10% to 15% of high-volume lung cancer centers. Ideal candidates for VATS lobectomy include patients who are properly staged with preserved pulmonary function and peripheral stage I tumors. Central tumors, tumors larger than 6 cm, andresection after neoadjuvant therapy for stage IIIA disease are relative contraindications. Absolute contraindications include those patients in whom a thoracoscopic technique cannot achieve complete resection (ie, chest wall or vascular invasion and the presence of ipsilateral mediastinal lymph node metastases).
Conclusion
Systematic mediastinal lymph node sampling or dissection should be a routine part of all anatomic resections for early-stage NSCLC to ensure the most accurate pathologic staging. Although some controversy surrounds the surgical management of stage I to II NSCLC, the best available data comparing lobectomy with sublobar resection suggest that sublobar resections should only be considered in select patients with small (< 2 cm) peripheral tumors or in those with impaired preoperative cardiopulmonary reserve. Although not proven with an appropriate randomized clinical trial, wedge resections have been associated with higher rates of local recurrence36; therefore, segmentectomy is often favored when sublobar resection is performed.
Dr. Lanuti: The standard of therapy for patients with early-stage lung cancer who are physically fit is lobectomy. However, if the tumor is smaller than 2 cm, the literature recommends considering sublobar resection for those patients, particularly those who have cardiopulmonary compromise.
Dr. Harpole: Surgeons in North America are interested in finding out the best way to treat patients with lesions smaller than 2 cm.
Dr. Lanuti: A small percentage of thoracic surgeons are doing VATS lobectomy. No prospective trials have compared VATS lobectomy with lobectomy, but based on some large studies, the approaches appear to be equivalent in terms of survival.
Dr. Lynch: My patients ask me, "Should I have this done thoroscopically or should I have a regular open resection?" I say, "Well, that's obviously the decision of the surgeon." But patients need to know the criteria for eligibility for VATS lobectomies.
Dr. Lanuti: Criteria for VATS lobectomy include peripheral lesions that are 3 cm or less. Larger lesions have been done, but the incision to remove the mass needs to be larger. Central lesions are not amenable to this technique. Does the patient need to be fit from a pulmonary standpoint? This assessment must always be performed, but severe pulmonary compromise is still a contraindication.
Dr. Lynch: What about prior pleural interventions? If you've had a car accident and had a chest tube, does that mean you can't undergo the procedure?
Dr. Lanuti: Previous chest surgery is not a contraindication.
Dr. Harpole: In fact, doing a lobectomy after a previous thoracotomy is a lot easier to do than a redo thoracotomy. It doesn't really matter whether you have a small incision or a large one, as long as the right cancer operation is done. Community surgeons need to know that a VATS wedge resection is not equivalent to an open lobectomy when it comes to being the right operation for a patient.