Lung cancer, which was first given status as a global epidemic in the 1950s, continues to be the leading cause of cancer-related deaths among both men and women.¹ Based on best available data, the worldwide incidence of lung cancer accounts for 1.2 million new cases and 1.1 million cancer deaths annually.² In 2007 in the United States alone, there were approximately 213,380 new cases of lung cancer and 160,390 lung cancer deaths.¹ It is the most common thoracic malignancy compared with esophageal cancer and mesothelioma, which account for approximately 12,000 and 3000 yearly cancer deaths, respectively. More deaths in the United States are due to lung cancer than to breast, prostate, and colorectal cancer combined.¹

The bulk of patients with lung cancer are divisible into two major groups based on treatment and prognosis: small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC). SCLC is the more aggressive form and usually has spread systemically by the time of diagnosis. The malignancy is characterized by a proliferation of small anaplastic cells. Because of its tendency to early metastasis, the cancer usually is not amenable to surgical resection, and hence, surgery does not play a primary role in the management of SCLC. It is, however, more responsive to systemic treatment with chemotherapy. The combination of etoposide and cisplatin remains the standard of care for both limited and extensive disease.³ Radiotherapy rather than surgery has been used for local tumor control. Recent innovations, including the addition of thoracic radiation to systemic chemotherapy protocols, increasing the intensity of thoracic radiation, neoadjuvant thoracic radiation, and prophylactic cranial irradiation,⁴ have produced some benefit in terms of prolonging disease-free intervals and survival.⁵ Untreated, the mean survival is 2–4 months. Median survival with treatment is between 18 and 36 months. Currently, SCLC accounts for 15–20% of new lung cancer cases per year in the United States.

NSCLC comprises of three major histopathologic subtypes, including squamous cell carcinoma, adenocarcinoma (including bronchioloaveolar cancer), and large cell (or undifferentiated/mixed) carcinoma. These cancers constitute approximately 80% of lung malignancies. They tend to spread more slowly than SCLC, and hence there are more opportunities for early intervention. Nevertheless, many patients with NSCLC have advanced disease at presentation. Surgery is the basis of treatment for early-stage NSCLC (stages I and II) and offers the best chance for cure. Stage III or IV lung cancers generally are treated palliatively with a variety of multimodality protocols, although selected patients with stage III disease have the potential for cure with resection depending on the degree of invasion of local structures and the extent of mediastinal nodal disease.⁶ A small proportion of lung cancers (<1%) exhibit no radiologic evidence of tumor. These cancers, termed occult, are diagnosed by screening bronchoscopy and sputum cytology.

All lung cancers share a common etiology in environmental or direct exposure to smoking tobacco. Cigarette smokers experience a 15- to 50-fold increased risk of developing lung cancer in comparison with lifetime nonsmokers. Smoking cessation among long-term smoking decreases lung cancer risk, with the diminution of risk proportional to years of smoking abstinence. More than 50% of lung cancer cases are currently diagnosed in former smokers. Nevertheless, 15–18% of lung cancers arise in individuals who have never smoked, and in this group, lung cancer represents the fifth most deadly cancer worldwide. There is an emerging association between bronchioloaveolar carcinoma and mutations of malignant cell surface receptors, specifically, the epithelial growth factor (EGF) receptors. These somatic mutations render tumors susceptible to treatment with specific inhibitors by tyrosine kinase cellular pathways. Also common to all lung cancer groups is the lack of an effective screening system for early detection, which poses a barrier to effective treatment.

EPIDEMIOLOGY

Despite an encouraging age-adjusted decline in lung cancer mortality rate in the 1990s, the absolute number of lung cancer deaths in the United States has increased dramatically since the 1950s because of a growing population of increasingly elderly persons (age > 70 years).⁷ Other characteristics of the lung cancer pool in the United States include a pronounced increase in this cancer in women, now claiming more lives than breast cancer, and a dramatic shift in the ratio of squamous to adenocarcinomatous histologic subtype in favor of adenocarcinoma.

The rise in adenocarcinoma of the lung correlates with changes in cigarette design and composition introduced by the tobacco industry after epidemiologic studies from the early 1950s demonstrated that cigarette smoking was strongly related to other forms of lung cancer. Cigarettes at the time were predominantly nonfiltered and contained high levels of both tar and nicotine. Such cigarettes were too irritating to permit deep inhalation into the lungs. Accordingly, lung cancers at the time were predominantly squamous cell and small cell carcinoma, both of which arise in central airways. On the other hand, individuals who smoked the new filtered low-tar, low-yield cigarettes compensated by inhaling reduced-yield cigarettes more deeply into the lungs. This resulted in a predominance of smoking-related adenocarcinoma of the lung.^8,9

The link between cigarette smoking and lung cancer was demonstrated epidemiologically in more than a dozen case-control studies of the early 1950s,¹⁰ followed by two prospective cohort studies from the United Kingdom.^11,12 The Surgeon General of the United States used these data in combination with established epidemiologic criteria of causality—consistency, strength, specificity, temporal relationship, and coherence of association between the disease and the disease-associated variable—to conclude in the 1964 Surgeon General's Report that "cigarette smoking is causally related to lung cancer in men." Because of the indisputable link between lung cancer and cigarette smoking, it is considered one of the most preventable forms of all human cancers. In the United States, 80% of cases can be attributed to smoking (90% men; 79% women). Direct evidence of this cause–effect relationship has been demonstrated using genetic amplification technique.¹³ Specifically, it has been demonstrated that a metabolite of benzo[]pyrene, a component of cigarette smoke, damages three specific loci on the p53 tumor suppressor gene. These loci have been found to be abnormal in approximately 60% of primary lung cancers.

Lung cancer susceptibility may be amplified by other environmental factors. Asbestos, radon, arsenic, ionizing radiation, haloethers, polycyclic aromatic hydrocarbons, nickel, family history, molecular genetic factors, presence of other benign lung disease (e.g., emphysema, chronic obstructive pulmonary disease, and interstitial lung disease), dietary factors (e.g., antioxidants and fat), and indirect (second-hand) exposure to cigarette smoke all have been implicated.

The lack of an effective screening program has done more to hamper clinical progress than all other impediments because resectability, prognosis, and survival are stage-dependent, and late-stage cancer remains incurable. The results of four lung cancer screening trials implemented in the 1970s have had a lasting impact on recommended practice guidelines for the treatment and management of lung cancer. Three of these trials were sponsored by the National Cancer Institute under the aegis of a program called the Cooperative Early Lung Cancer Detection Program: The Mayo Lung Project,¹⁴ the Memorial Sloan Kettering Lung Project,¹⁵ and the Johns Hopkins Lung Project.¹⁶ A fourth trial was conducted in eastern Europe: The Czechoslovakia Study on Lung Cancer Screening.¹⁷ The screening method employed in all these studies consisted of plain-film chest x-ray with sputum sampling for cytology.

On the basis of these clinical trials, the American Cancer Society was unable to recommend lung cancer screening as the standard of care because none of the randomized studies demonstrated a statistically significant correlation between lung cancer screening and mortality reduction.¹⁸ All the positive outcome variables used in these trials, save for "mortality," were considered on the basis of inductive reasoning to be hypothetically confounded by statistical biases related to early detection trials. These potential confounding influences included lead-time bias, length bias, selection bias, and overdiagnosis bias. However, these biases may have little impact on lung cancer survival because of the aggressive nature of lung cancer compared with other malignancies, such as prostate cancer, which has a more indolent course. Furthermore, it is possible that the impact of screening programs in terms of obtaining a positive result requires long-term follow-up (>5 years) for other than incidence cases. As a consequence, other variables such as stage distribution, resectability, survival, and fatality, although statistically significant, were deemed subordinate to mortality, which failed to show a statistically significant correlation in any of the clinical trials. These methodologic issues and recent improvements in CT technology, as well as the implementation of parallel screening programs for other large-scale solid tumors (e.g., breast and colon cancer), have fueled efforts by the public health community to revisit lung cancer screening as a recommended practice.

HISTOPATHOLOGIC CLASSIFICATION

The histologic subtypes of lung cancer are detailed in Table 60-1, along with the spectrum of other lung and pleural tumors, both malignant and benign. This classification system was developed by the World Health Organization in 1967 and has been revised several times. The most recent revision was published in 1999.¹⁹

Adapted from Travis et al.¹⁹ (WHO, 1999).

Squamous and small cell lung cancers tend to arise in the central airways, whereas adenocarcinoma and large cell lung cancers tend to locate peripherally. Small cell lung cancers arise from neuroendocrine cells that are distributed in small numbers in the normal epithelium. There are four major types of neuroendocrine tumors: small cell neuroendocrine, large cell neuroendocrine, and typical and atypical carcinoids. Typical carcinoids grow slowly and rarely spread beyond the lungs. Atypical carcinoids, which are rare, exhibit more rapid growth and are more likely to spread to other organs. The location of the carcinoid tumor (i.e., central, peripheral, or endobronchial) dictates the treatment approach; however, prognosis following resection for this tumor depends on an R₀ resection (i.e., gross and microscopic complete tumor removal) and regional node dissection.

Squamous cell carcinomas arise from metaplastic squamous cells because squamous cells are absent from the normal epithelium. Adenocarcinomas arise from Clara cells or type 2 pneumocytes, the precursors of bronchioles and alveoli, respectively. Large cell cancers consist of poorly differentiated cells. Mucoepidermoid tumors arise from tracheobronchial mucus glands and have similar cellular features to mucoepidermoid tumors that originate in the salivary glands.

Identifying the specific histologic subtype confirms the diagnosis and can provide important cues in certain clinical situations, but the more important distinction in terms of management and treatment is between SCLC and NSCLC.

ANATOMY OF THE LUNG AND TRACHEOBRONCHIAL TREE

The lung has two main lobes (left and right) (Fig. 60-1). The right lung is marginally larger because the left lung accommodates the heart by having only 8 segments compared with the right lung, which has 10 segments. Each lung has at least one fissure that divides the lung into smaller lobes. The left lung is divided in two by a single horizontal fissure that creates an upper and lower lobe. The right lung has two fissures, one horizontal and one oblique. These fissures delineate three lobes: upper, middle, and lower. A normal anatomic variant includes the presence of an azygos lobe (see Fig. 60-1, inset), which is usually found at the apex of the right lung. This small variant lobe is separated from the upper lobe by a deep fissure-like groove that cradles the azygos vein. The lobes of the left and right lung, in turn, are divided into segments representing areas of lung served by different bronchioles, as shown in Fig. 60-2. This figure also shows the intimate relationship between the lungs and tracheobronchial tree. The trachea lies anterior to the esophagus (not shown). At the bifurcation of the trachea, or carina, the left and right mainstem bronchi branch off, and each branch enters the hilus of its respective lung. These, in turn, divide into progressively smaller airways, called bronchioles, that form a rootlike network that extends through the spongelike tissues of the lung. The exterior layer of the bronchi is composed of cartilage with rings of smooth muscle that permit the bronchi to expand and retract on inspiration and expiration. The cartilaginous segments become more irregular at the distal end of this network, and there are none on the bronchioles.

CLINICAL PRESENTATION NSCLC

Patients are candidates for surgical resection when there is a diagnosis or reasonable probability of NSCLC. Despite the disparate histology, these patients share similar prognosis and are managed with a unique staging system discussed below. Primary NSCLC tumors are often peripheral and grow more slowly than SCLC. Symptomatic manifestations of local disease in NSCLC include cough, hemoptysis, chest pain, dyspnea, wheezing, and pneumonia. Symptoms of locally advanced disease include hoarseness, phrenic nerve paralysis, dysphagia, stridor, superior vena cava syndrome, pleural effusion, pericardial effusion, Pancoast syndrome, evidence of lymphangitic spread, and cancer cachexia. Manifestations of extrathoracic spread include brain metastases, bone metastases, and spread to liver, adrenal glands, and intraabdominal lymph nodes. At clinical presentation, 60–75% of patients have cough, weight loss, or dyspnea. Hemoptysis, chest or bone pain, fever, and weakness occur slightly less often.²⁰ The physical examination foretells advanced disease if, among other signs, there is evidence of lymphadenopathy in the supraclavicular or cervical regions, percussion dullness suggesting a pleural effusion, or neck vein distention from superior vena cava obstruction²⁰ (Table 60-2).

TNM STAGING SYSTEM FOR LUNG CANCER

All cancers follow a predictable progression from primary tumor to local invasion to metastasis by means of lymphatic or hematogenous dissemination or both. This is particularly relevant to thoracic malignancies because the lymphatic system in the lungs drains unidirectionally along the bronchi to the hilum and then to the mediastinal nodes. This lends greater accuracy to staging and better prognostic correlation for cancer resections by anatomic lobar dissection with hilar and mediastinal node removal. The confounding difficulty is that skip metastases can occur in up to 25% of patients. The current staging system used for NSCLC lung cancers²¹ defines the extent of disease based on the size of the tumor (T), the extent of nodal involvement (N) (Fig. 60-3), and the presence of distant metastases (M). The TNM concept was proposed originally by Denoix in the 1940s²² and later adopted by the American Joint Committee on Cancer and the International Union Against Cancer in 1986.²³ The system has been revised several times in recent years. The current version (edition 6) was revised by the International Association for the Study of Lung Cancer (IASLC) in 1999.²⁴ A new version (edition 7) still under discussion by the IASLC proposes several changes in TNM descriptors that will affect existing treatment algorithms but brings clinical stages more in line with prognosis²¹ (Table 60-3) (see Fig. 60-5).

Adapted from reference 21.

Good staging yields the ability to stratify patients into homogeneous groups by prognosis, guides therapeutic decision making, and permits more accurate cross-study comparison of treatment strategies and outcomes. Even for patients not determined to be suitable for surgery, staging is an essential guide to treatment and prognosis for other treatment modalities such as chemotherapy or radiotherapy.

The TNM classification of malignant tumors consists of eight stages: IA, IB, IIA, IIB, IIIA, IIIB, IV, and 0, which is applicable to patients with carcinoma in situ. The relative location of tumor based on disease stage (I-III) is shown in Figure 60-4. Staging is not relevant for occult lung cancer. Stage I and II (early) lung cancer is amenable to surgical resection (see Chaps. 62, 63, 64, and 65). Patients with stage III tumors (i.e., locally advanced) may be potential candidates for surgery depending on the characteristics of local invasion. If the tumor has infiltrated the superior vena cava or the carina, resection may be warranted under certain conditions. However, if other mediastinal structures such as the great vessels, the esophagus, or the heart are involved, or if there is nodal metastasis at undesirable levels, resection is often precluded (see Chap. 70). R₀ resection is the goal of all stage I and II resections; R₁ (microscopic disease is left) or R₂ resections (i.e., gross tumor is left) rarely lead to cure of NSCLC. Surgical resection for stage IV NSCLC (i.e., extensive metastatic spread) is usually contraindicated except in the uncommon scenario of surgery for oligometastatic surgery, whereby resection with curative intent of proven solitary pulmonary metastases may be included either prior to or after lung resection (see Chap. 71).

Occult lung cancer (i.e., positive sputum cytology but radiologically negative findings) is extremely rare, comprising less than 1% of all lung cancers (see Chap. 53). The diagnosis is made by screening bronchoscopy used to detect endobronchial lesions with cytologic brushing and washing of individual lobar bronchi. Surgical treatments for occult disease include lung-sparing procedures, that is, bronchoplastic techniques (see Chap. 66). Other surgical procedures offered to patients in poor medical condition to provide palliative relief may be beneficial, for example, restoration of airway patency through bronchoscopic debridement of tumor, photodynamic therapy, airway stenting, or brachytherapy.

According to the current edition (sixth) of the TNM Classification of Malignant Tumors:

Tumor (T): T₁ denotes a solitary primary tumor that is 3 cm or less in diameter surrounded by lung or visceral pleura and extends no more proximal than a lobar bronchus. A T₂ tumor is greater than 3 cm in diameter and invades the visceral but not parietal pleura. All T₃ tumors are locally invasive. Invasion may extend to the parietal pleura, chest wall, diaphragm, mediastinal pleura, parietal pericardium, or mainstem bronchus if it is less than 2 cm from the carina. There also may be atelectasis or pneumonitis involving the entire lung. T₄ tumors may invade the carina, trachea, mediastinum, great vessels, heart, or esophagus. Additionally, if there is malignant pleural or pericardial effusion, the T₄ status applies irrespective of other features. The T₄ designation is also applied automatically if satellite tumor nodules are found in the same lobe as the primary lesion. However, ipsilateral pleural effusions designated T₄ may not be a pathologic T₄ tumor because thoracoscopy with pleural biopsy may reveal a benign effusion secondary to lobar collapse. Furthermore, clinical T₄ tumors from a secondary nodule within the same lobe demonstrate better prognosis than other T₄ tumors. Therefore, patients with secondary tumors in the same lobe as the primary should be offered a resection in the absence of distant metastases.²⁵

Node (N): N₀ indicates the presence of a primary lesion without metastasis to regional lymph nodes. N₁ denotes metastatic spread of cancer to one or any of the following lymph node stations: ipsilateral peribronchial lymph nodes, ipsilateral hilar nodes, or both, and intrapulmonary nodes involved by direct extension of the primary tumor (lymph nodes with double-digit numbering). The N₂ designation indicates metastatic spread to lymph nodes of the ipsilateral mediastinum, subcarinal nodes (single-digit nodes), or both. N₃ is metastasis to contralateral mediastinal or hilar lymph nodes or ipsi- or contralateral scalene or supraclavicular lymph nodes. Documentation of the pattern of thoracic of lymph node drainage preceded the development of the TNM staging system.²⁶

Metastasis (M): M₀ denotes the absence of distant metastasis. M₁ indicates the presence of a distant metastasis.

DIAGNOSTIC STUDIES AND STAGING METHODS

Noninvasive radiologic evaluation begins with a posteroanterior chest x-ray and chest CT scan through the adrenal glands to assess disease and determine the size of the lesion, involvement of surrounding structures, and overall resectability (Table 60-4). The sensitivity of CT scanning for detecting lung cancer is between 50% and 80%. MRI, once thought to hold promise as a diagnostic tool, is usually reserved for advanced anatomic problem solving when the results of CT scanning are ambiguous.²⁷ PET scanning has high sensitivity (range 79–95%) for detecting lung cancers, in particular distant metastatic foci of disease, but low specificity. Integrated CT/PET is both sensitive and specific, yielding 98% correct tumor staging compared with final histopathologic staging. These technologies are expanding, but the instruments are expensive and not as widely available as CT.²⁸ Although CT scanning remains the mainstay of clinical staging, CT scanning alone is inaccurate because it has a limited ability to determine mediastinal nodal involvement, chest wall invasion, mediastinal invasion, and malignancy of pleural effusions.^28,29

Abbreviations: CXR = chest x-ray; CT = computed tomography; PET = positron-emission tomography; CT/PET = integrated computed tomography and positron-emission tomography; EUS = endoscopic ultrasound; EUS/FNA = endoscopic ultrasound with fine-needle aspiration; EBUS = endobronchial ultrasound; EBUS/FNA = combined endobronchial ultrasound and fine-needle aspiration.

A number of staging and diagnostic procedures have evolved using the bronchoscope. Invasive bronchoscopic procedures include transbronchial needle aspiration, endobronchial ultrasound biopsy, endoesophageal ultrasound fine-needle aspiration, and transbronchial and transesophageal needle biopsy. Although radiologic reconstructions of the bronchi and trachea are useful adjuncts to operative planning, bronchoscopy is recommended before every surgical resection as an aid to identifying anatomic abnormalities that might interfere with surgical resection and to confirm the accuracy of noninvasive study by determining the proximal extent of the tumor (i.e., distance from the carina). It is also valuable in identifying occult synchronous disease or anatomic variants of normal.

Invasive preresectional surgical staging, pioneered by the Thoracic Surgical Group of Toronto, is routine practice at many centers.^30,31 Mediastinal staging via mediastinoscopy was conceived originally to spare the morbidity of an exploratory thoracotomy.³⁰ It is used presently to document the anatomic extent of the lung cancer before surgical resection, as a guide to therapeutic decision making, and for restaging lung cancer³² (see Chap. 61). Selective lymph node sampling via cervical mediastinoscopy, for example, helps to identify patients with minimal N₂ disease (stage III) for neoadjuvant chemotherapy or multimodality protocols (e.g., chemotherapy, radiotherapy, and surgery). Many predict that the role of mediastinal staging in lung cancer will expand with the increasing availability and application of new diagnostic and therapeutic techniques, such as tissue biopsy for biomarker staging³³ and tissue typing for oncogenic mutations³⁴ and drug-sensitivity testing.³⁵

Thoracoscopy is usually performed intraoperatively in patients who are not suspected of having mediastinal lung cancer for confirmation of lymph node histology and guidance for postoperative chemotherapy.

Meta-analysis has been used to demonstrate the superiority of pathologically staged lung cancer. Survival in clinically staged IB patients (T₂N₀ disease) was only 75% at 1 year and 40% at 5 years.³⁶ In contrast, pathologically staged T₂N₀ disease yielded 90% survival at 1 year and 60% at 5 years,³⁶ highlighting the limitations of current CT technology. Attempts to use noninvasive techniques to replace mediastinal staging by mediastinoscopy results in a 10% error regardless of the noninvasive technique used.

Lesions located in specific lobes drain preferentially and predictably to specific nodal groups. Cervical mediastinoscopy is used to access nodes at the paratracheal levels (2R, 2L, 4R, 4L, 10R, and 10L) in addition to the subcarinal nodes at level 7. Anterior mediastinoscopy (also called parasternal mediastinoscopy) accesses lymph nodes at levels 5 and 6.

Thoracoscopy of the right hemithorax accesses the right paratracheal nodes (level 4R), inferior pulmonary ligament nodes (level 9), and subcarinal nodes (level 7). In the left hemithorax, thoracoscopy is used to access the aortopulmonary window nodes (levels 5 and 6) and the posterior hilar nodes.

Mediastinal staging by mediastinoscopy is a safe and effective technique that provides important histologic information with minimal morbidity. It can be performed on an outpatient basis. It is currently indicated for patients with lung cancer suspected to have spread to the mediastinal nodes to provide accurate lymph node staging and histologic confirmation, which is used to guide therapy. Patients not suspected of having mediastinal lymph node involvement may be staged with preoperative CT/PET and intraoperative thoracoscopic staging to confirm the absence of mediastinal malignancy. The drawbacks to preresectional staging with mediastinoscopy are twofold: It requires general anesthesia and delays the surgical resection.

PREOPERATIVE ASSESSMENT

Preoperatively, all patients must undergo pulmonary function testing to assess their overall fitness for surgery and to predict postoperative outcome. These studies include forced vital capacity (FVC), forced expiratory volume in 1 second (FEV₁), and diffusion capacity of carbon monoxide (DLCO). The FEV₁ then is used to calculate the predicted postoperative FEV₁. A ppoFEV₁ of 0.8–1.0 L or more is desired for lung resection of any kind. This value is the product of FEV₁ and the percent of lung parenchyma that will remain after resection. Residual lung parenchyma can be estimated by the percent of remaining pulmonary lobes (x of 5) or remaining anatomic pulmonary segments (x of 18) or estimated percent of remaining ventilation or perfusion based on the / scan.

STANDARD PULMONARY RESECTIONS

Depending on the size, location, and extent, the lesion may be resected using an anatomic (e.g., lobectomy, pneumonectomy, bilobectomy, or sleeve lobectomy) or nonanatomic (e.g., segmentectomy, wedge resection, precision cautery, or metastasectomy) resection. Lobectomy is the standard anatomic resection for NSCLC when the lesion occurs within the boundary of an anatomic lobe (see Chap. 62). The most common technique begins with a thoracotomy incision followed by hilar division of the lobar pulmonary arteries, draining pulmonary lobar veins, and associated bronchus. This is followed by meticulous dissection of hilar lymph nodes and lymphatics, which then are removed en bloc with the specimen lobe as a single unit. Alternatively, a sternotomy incision may be used for access to bilateral pleural spaces, mediastinum, and anterior hila in the presence of bilateral pulmonary nodules. This approach and muscle-sparing thoracotomies are associated with decreased postoperative pain. They may be used for early stage I or II NSCLC.

MINIMALLY INVASIVE SURGICAL TECHNIQUES

Less invasive pulmonary resections using video-assisted thoracic surgery technique, thoracoscopy, and other minimally invasive surgical techniques have improved the morbidity of surgical resection and afford surgery in patients with compromised pulmonary function (Table 60-4). The benefits of these techniques include shorter postoperative stay, reduced narcotic requirements for postoperative pain management, reduced shoulder dysfunction compared with patients undergoing thoracotomy, and greater patient satisfaction.

LUNG CANCER SURGERY IN THE ELDERLY

More than half of all cancers are diagnosed in patients 65 years or older, yet the risk of surgery is greatest in this group. Paradoxically, elderly patients also have a higher rate of early-stage cancer detection compared with younger adults. Lung-sparing resections using minimally invasive techniques, such as limited thoracoscopic wedge resection, are increasingly being offered to elderly patients (age > 65 years) who have impaired or diminished respiratory function along with other comorbid conditions that make them unfit for pneumonectomy or lobectomy.³⁷ Postoperatively, pulmonary complications have been identified as the major cause of morbidity and mortality after surgery in elderly patients. It is reasonable to assume that a reduction in pulmonary complications through the use of less invasive lung-sparing procedures will lead to a decrease in morbidity and mortality in the elderly age group.

FIVE-YEAR SURVIVAL IN RESECTABLE NSCLC

The importance and value of accurate staging are evident in studies that report cumulative 5-year survival rates for NSCLC. Broadly estimated, 5-year survival, irrespective of stage, is 10–15%; this low survival rate has improved only minimally despite multiple new modalities of treatment over the last 50 years. For stage I and II disease combined, it is 50%. Examined individually, these reports demonstrate considerable variability. Reviews published before the 1997 revision of the TNM staging system differ from reports published thereafter, reflecting the expansion of stages I and II (i.e., stage IAB and IIAB) and the reassignment of certain stage III patients to stage IIB based on prognosis. A report published in 1995 estimated that the 5-year cumulative survival rate for patients with stage I disease was 64.6%, and for those with stage II disease it was 41.2%.³⁸Naruke and colleagues³⁹ and Mountain³⁶ independently evaluated 5-year survival in stage IA+B and stage IIA+B disease using the 1997 revised IASLC. A summary of their findings reveals that the 5-year survival in stage IA+B disease combined is between 67% and 75%, whereas the limit of survival in stage IIB disease was no better than 38%. This reduced survival for stage II disease reflects the adverse effect of nodal metastasis, which portends aggressive tumor biology and behavior.

Goldstraw and colleagues recently have compared clinical and pathologic 5-year survival between the sixth and seventh editions of the TNM Classification of Malignant Tumors, which is under review²¹ (Fig. 60-5). Among the proposed changes to TNM descriptors, recently reviewed by the International Association for the Study of Lung Cancer's Lung Cancer Study Group, additional cutoffs have been recommended for tumor size. Tumors larger than 7 cm will move from T₂ to T₃. Additionally, pleural effusion will be reclassified as an M descriptor, T_2bN₀M₀ patients will move from stage IIB to stage IIA, and T₄N_0-1M₀ patients will move from stage IIIB to stage IIIA. These alterations will align prognosis and TNM stage more closely; in certain subsets, it will help to align treatment with stage.

LUNG CANCER SCREENING REVISITED

Currently, no public policy organization recommends screening for lung cancer as the standard of care among high-risk individuals. The public health community has focused its strategy on controlling the lung cancer epidemic exclusively on prevention by promoting cigarette smoking cessation and preventing smoking initiation among children and adolescents. Aside from the fluctuating efficacy of this approach, this strategy ignores the risk of lung cancer in former smokers (one in four adults in the United States are now former smokers) and lifetime nonsmokers (15% of lung cancers arise in lifetime nonsmokers). Moreover, not all countries have the political will or funds to conduct smoking-cessation programs. The large randomized controlled trials of the 1970s and 1980s^14–17 unanimously concluded that chest x-ray and sputum cytology were too insensitive to detect early-stage lesions or to have any measurable impact on mortality. Renewed effort to implement a cost-effective early detection program for lung cancer has been spurred by recent technologic advances across several disciplines, most particularly in the field of radiographic imaging. CT technology, the current diagnostic standard, is rapidly advancing; however, the pace of this technology makes randomized clinical trials with proven endpoints (e.g., mortality reduction secondary to screening) impractical. The technology is changing so fast that studies are obsolete by the time that sufficient accrual is attained. Consequently, although a number of nonrandomized studies have been performed in different countries demonstrating the utility of low-dose CT for detecting lung cancers at early, preinvasive stages before they are evident clinically, these studies have generated controversy in the public health community because they do not follow established concepts of a screening study.^40–42

A second problem relates to cost. Based on the experience from the Mayo Clinic CT screening study, one analysis estimated that the cost of lung cancer screening may be as high as $116,300 per quality-adjusted life-year.⁴³ However, this conclusion has been challenged. Data from the International Early Lung Cancer Action Project and recent studies from Italy have been used to conclude that the average cost of low-dose CT is US$2500 per year of life saved.⁴⁴ This cost may be a justifiable expense to screen a population of at-risk individuals given the expense and ineffectiveness of treatment for advanced-stage disease. In comparisons with other screening techniques, it has been suggested that the detection rate of lung cancer with low-dose CT is somewhat greater than with mammographic screening for breast cancer.⁴⁵

Certain public agencies are beginning to move on the issue of lung cancer screening, whereas other entities remain staunchly opposed.^46,47 The U.S. Preventive Services Task Force recently issued an indeterminate (I) grade summary recommendation concluding that "the evidence is insufficient to recommend for or against screening asymptomatic persons for lung cancer with either low-dose computerized tomography, chest x-ray, sputum cytology, or a combination of these tests." Reasons for this cautionary move from "not recommended" to "indeterminate" include the potential for overdiagnosis,⁴⁸ because low-dose CT scanning is associated with a greater number of false-positive results,⁴⁹ increased radiation exposure, and increased costs compared with chest x-ray. This practice could lead to undue worry and concern on the part of patients who undergo invasive diagnostic procedures only to be proved to have benign findings. On the basis of this and other similar advisories, practitioners are now counseled to inform their patients on the availability of lung cancer screening modalities in view of the potential risks, as well as the pros and cons of the present art of lung cancer screening, as described earlier. Patients are then free to make their own informed decision. This topic is relevant from a surgical perspective because surgery remains the best chance for cure in patients with early stages of NSCLC, and identification of early-stage disease is currently the only avenue for mitigation of this worldwide epidemic.

SUMMARY

The importance of having a universally adopted international lung cancer staging system, as we do, is most evident when it comes to evaluating multimodality treatment protocols consisting of combinations of surgery, chemotherapy, radiotherapy, and other innovative treatments. Although surgery alone has been the mainstay of treatment for early-stage NSCLC, there is now compelling evidence from randomized trials that adjuvant chemotherapy improves survival following resection in stage II and IIIA NSCLC, as well as evidence that chemotherapy is effective in stage IB tumors greater than 4 cm in diameter. Meaningful analysis of the numerous multimodality treatment protocols would be impossible without a universal staging system. This reliance is increased with the continuing expansion of new and divergent clinical methods to permit earlier lung cancer detection through screening, when surgery has the greatest chance for cure, or to tailor multimodality treatment protocols capable of preventing recurrence and improving survival.

EDITOR'S COMMENT

This overview emphasizes the new WHO classification as well as the new TNM staging system. With the emergence of bronchioloaveolar carcinoma (BAC) and increased use of adjuvant chemotherapy in non-small-cell lung cancer (NCSLC), the importance of pretreatment staging cannot be overemphasized. Finally, a discussion of lung cancer in the elderly and lung cancer screening, two current controversial topics, is offered.

–MJK

REFERENCES