According to American Cancer Society statistics, there will be 232,270 new cases of lung cancer in 2008, accounting for approximately 15% of cancer diagnoses. The likelihood of developing lung cancer is 1 in 2500 in men younger than 39 years of age and 1 in 15 in men between the ages of 60 and 79 years.

The vast majority of patients with lung cancer will still die of the disease. It is estimated that 166,280 patients will die of lung cancer in 2008, accounting for 31% of all cancer deaths in men and 26% of cancer deaths in women. Among women, the rate of deaths from lung cancer remains high. Since 1987, more women have died from lung cancer than from breast cancer. Despite an enormous effort to find effective therapies for lung cancer, the overall 5-year survival remains only 15%. The high mortality rate and low 5-year survival rate are largely the result of an inability to diagnose lung cancer at an early stage, when it is still potentially curable. Only 15% of lung cancers are diagnosed at this early stage. The uniformly poor results obtained with current therapies are a strong argument for the inclusion of all possible patients into clinical trials.

Risk Factors

Tobacco Smoking

Smoking is by far the most common cause of lung cancer. Approximately 85% to 90% of lung cancer cases can be attributed directly to tobacco smoking. The risk of lung cancer is directly related to the length of time a person smokes, the number of cigarettes smoked per day, the age at which a person starts smoking, and the amount of tar contained in the cigarettes. The risk of developing lung cancer is 9- to 10-fold higher in an average smoker and up to 25-fold higher for a heavy smoker. However, not all tobacco smokers develop lung cancer. This may be attributable to differences in an individual’s inherited predisposition to cancer development. There is also evidence that nonsmokers exposed to tobacco smoke (passive smokers) have an increased risk of developing lung cancer. It has been estimated that up to 25% of lung cancers in nonsmokers are caused by passive smoking. The risk of lung cancer declines steadily once a person stops smoking, but it takes up to 15 to 20 years for the risk to return to the level of nonsmokers. However, it never reaches this level if they had smoked two packs per day or more.

Other Risk Factors

Other factors reported to cause lung cancer include occupational exposure to arsenic, asbestos, nickel, uranium, chromium, silica, beryllium, and diesel exhaust. Dietary deficiency of vitamins A, C, and E, retinoids, carotenoids, and selenium, and air pollution, lung scars, and oncogenes have all been implicated in the causation of lung cancer. Asbestos is the most common occupational cause of lung cancer and increases the risk of lung cancer fivefold. Tobacco smoking and asbestos exposure act synergistically, and the risk of lung cancer in smokers who have been exposed to asbestos becomes as high as 80-fold to 90-fold higher than the risk in nonsmokers. Radon, produced by the decay of uranium, has a well-established association with lung cancer in underground miners. However, the association between lung cancer and residential exposure to radon is less well established.

Most patients with lung cancer present with advanced-stage disease, when the outcome is dismal. Therefore, there has been an interest in screening high-risk people to detect lung cancers when they are smaller and presumably at earlier and more curable stages. Earlier trials using chest radiography, cytologic examination of the sputum, or both, however, failed to decrease the disease-specific mortality. These trials have been criticized for their study design, statistical analysis, contamination, and older forms of technology. With the development of newer imaging technology, specifically low-dose spiral CT scans, there has been a new interest in screening for lung cancer. The International Early Lung Cancer Action Program screened 31,567 asymptomatic patients, with 27,456 repeat screenings and found lung cancer in 484 participants. Of these, 85% had stage I disease with an estimated 10-year survival rate of 88%. Although intriguing, this study has been criticized on methodological grounds. Another longitudinal study of 3246 asymptomatic patients from three centers found 144 lung cancers, compared to 44.5 expected cases. There was no evidence of a decline in the number of diagnoses of advanced lung cancers.

In addition to the overdiagnosis of small tumors detected during screening, which could otherwise remain silent until the patient dies from other causes, screening for lung cancer with spiral CT scan could potentially cause harm secondary to morbidity and mortality associated with thoracotomy, since lung tumors appear as noncalcified nodules and only a small fraction of those nodules actually represent lung cancer. The ongoing National Lung Cancer Screening Trial is carefully studying the role of CT scans in the screening of lung cancer in patients with a strong history of tobacco smoking. The issues of lead-time bias and additional morbidity and mortality are particularly relevant in the population of older adults in whom competing causes of death and multiple comorbidities are frequently encountered. At the present time, we do not believe there is enough evidence to recommend CT screening for lung cancer outside the context of clinical trial.

With better understanding of the molecular pathology of lung cancer, an interesting area of research is to examine the utility of molecular and genetic biomarkers in screening high-risk people. Because cancer is a multistep process, identification of these biomarkers in the sputum of high-risk individuals may help to identify the early clonal phase of progression of lung cancer. If this becomes reality, then it may enable detection of cancers earlier than with spiral CT. It could also be complementary to spiral CT screening.

Based on the light microscopic features, lung cancers are divided into two major groups: small-cell lung cancer (SCLC) and non–small-cell lung cancer (NSCLC). The later include squamous cell carcinoma, adenocarcinoma, and large-cell carcinoma (Table 97-1). Bronchoalveolar carcinoma is subclassified under adenocarcinoma. The distinction between small- and non–small-cell histology is important clinically, because small-cell histology is more aggressive and more responsive to chemotherapy and radiation.

In the past, the most common histology was squamous cell carcinoma, which accounted for 40% of lung cancers. However, now in the United States only 20% to 25% of lung cancers are squamous cell cancers. Squamous cell carcinoma is the classic lung cancer—central in location, endobronchial in nature, sometimes with central cavitation, and commonly associated with lobar collapse, obstructive pneumonia, or hemoptysis, with late development of distant metastases. On the other hand, adenocarcinomas, which in earlier times constituted 20% to 25% of all lung cancers, are now the most prevalent histology, accounting for almost 40% of all lung cancers. This increase in prevalence has been greater in North America than in Europe, where squamous cell cancer is still the most common type. Changes in smoking habits and other environmental factors are thought to account for this shift in histology.

Adenocarcinoma is not as strongly associated with smoking as other types of carcinomas and is the most common type in nonsmokers. It is characterized by the early development of metastases, even when tumor still appears to be a solitary lesion. Adenocarcinoma is most frequently a peripheral lung lesion and is commonly accompanied by a malignant pleural effusion.

The incidence of squamous cell carcinoma rises with increasing age, while that of adenocarcinoma falls. This, in combination with the fact that elderly patients seem to have more localized disease at presentation, correlates with a greater likelihood that older patients will have resectable and thus potentially curable disease.

Bronchoalveolar carcinomas and large-cell carcinomas are less common. The former is either unifocal or multifocal, which can mimic pneumonia. It may be detected in the setting of prior fibrotic lung disease, such as idiopathic pulmonary fibrosis, scleroderma, or asbestosis. Large-cell carcinomas account for approximately 10% of all lung cancers. The lesions are usually peripheral and sometime cavitate because of necrosis. Small-cell carcinomas account for approximately 20% of all lung cancers. Among all the lung cancer types, they have the strongest association with smoking. These commonly form central masses and grow rapidly, with a tendency to early metastasis. Rarely, tumors contain two histologies.

The initial clinical presentation of lung cancer patients is variable. Patients may seek medical attention for symptoms related to the primary tumor, mediastinal disease, distant metastases, or paraneoplastic syndromes (Table 97-2). In advanced cases, patients may present with systemic signs of anorexia, weight loss, fatigue, and weakness. Symptoms related to the primary tumor depend upon the location of the tumor: central or peripheral. Central tumors produce cough, dyspnea, hoarseness, stridor, hemoptysis, and postobstructive pneumonia. Superior sulcus tumors may produce shoulder pain, arm pain, or brachial plexopathy, or Horner syndrome. Peripheral lesions may cause pleuritic chest pain suggesting pleural involvement. Symptoms related to mediastinal disease include hoarseness caused by the involvement of left recurrent laryngeal nerve with left-sided tumors and obstruction of the superior vena cava with right-sided tumors or associated lymphadenopathy. Less common features include dysphagia caused by esophageal obstruction and hypotension caused by pericardial tamponade. Direct involvement of the myocardium is rare, although pericardial effusions are sometimes seen and may produce severe dyspnea and refractory hypotension. Phrenic nerve involvement leads to paralysis of the ipsilateral diaphragm, most commonly seen as elevation of the affected diaphragm on a plain chest radiograph.

Metastases from lung cancer are common. Sites of spread include the brain, pleural cavity, bone, liver, adrenal glands, and contralateral lung. Because more than half of patients have metastases at presentation, initial symptoms related to a metastatic site are frequent, particularly in patients with adenocarcinoma or small-cell carcinoma.

Although lung cancer is associated with a variety of paraneoplastic syndromes, presentation with symptoms related to a paraneoplastic syndrome is uncommon. Apart from hypercalcemia and hypertrophic pulmonary osteoarthropathy that are associated with squamous cell carcinoma, all others are linked to SCLCs. These include the syndrome of inappropriate secretion of antidiuretic hormone (SIADH), Cushing syndrome, and various neurological syndromes such as Eaton–Lambert reverse myasthenia syndrome, encephalitis, subacute sensory neuropathy, opsoclonus and myoclonus, cerebellar degeneration, and limbic degeneration. Hypercoagulability resulting in deep venous thrombosis is common in advanced lung cancers, particularly adenocarcinoma. Occasionally, lung cancers are diagnosed during work-up of patients with dermatomyositis or membranous glomerulonephritis.

An accurate diagnosis with proper distinction between SCLC and the various non-small-cell types is essential before any further staging work-up can be performed. A change in the usual symptoms of cough and dyspnea in a chronic smoker is often an indication for a chest x-ray, which if abnormal, leads to further evaluation. Because benign and other malignant conditions can mimic lung cancer on radiologic studies, histological confirmation of the diagnosis is essential. Depending on the location of the tumor, this can be achieved through examination of sputum cytology, fiberoptic bronchoscopy, or transthoracic biopsy.

Sputum cytology is a cost-effective and easy way of reaching a diagnosis. The yield is high for endobronchial lesions such as small-cell tumors and squamous cell carcinomas, but it is low for peripheral lesions of adenocarcinoma and large-cell carcinoma. The false-positive rate for cytology is less than 1%, but the false-negative rate is 41% to 43%. A cytological specimen that is positive for malignancy is accurate in approximately 90% of cases, but the distinction between individual histologic types is not as accurate, and discordant results are often seen between cytology and histology bronchoscopy and needle-biopsy specimens.

In contrast, fiberoptic bronchoscopy provides direct visualization of the bronchial tree, and diagnostic material can be obtained with bronchial brushings, washings, and biopsy of the directly visualized tumor, or with a transbronchial biopsy. Fiberoptic bronchoscopy is well tolerated by elderly patients, and age alone should not exclude an appropriate diagnostic evaluation. For tumors not visualized, the yield for washing and brushing is approximately 75% in central lesions and 55% in peripheral lesions. Therefore, the yield in small-cell and squamous cell carcinomas is higher than in adenocarcinomas and large-cell carcinomas.

Percutaneous fine-needle aspiration of a lung lesion gives a high yield (more than 85%), with a high accuracy for histological subtype. This technique is very useful to establish a diagnosis for peripheral lesions and in patients who are not surgical candidates for an open thoracotomy. One deterrent for performing percutaneous fine-needle aspiration is the risk of inducing iatrogenic pneumothorax; however, chest tube placement is required only in approximately 5% of patients. The same technique is used in obtaining tissue from metastatic sites such as supraclavicular lymph nodes, adrenal glands, and liver. In patients with a classic solitary lesion that is surgically resectable, many times a diagnosis is made after a thoracotomy. This is because the management of the lesion is unlikely to be changed if a percutaneous biopsy is performed first. In patients with a pleural effusion, either cytological examination of the pleural fluid or, if that is inconclusive, thoracoscopic biopsy of pleural lesions is the method of obtaining a biopsy material. An additional method of obtaining a tissue is mediastinoscopy.

Staging