DEFINITION

Lung cancer is a major public health problem. In the United States, 31% of cancer deaths in men and 26% of cancer deaths in women are secondary to lung cancer. Efforts at early detection and treatment have been frustrating, and hence the overall prognosis remains poor. Just over one in eight lung cancer patients will be living 5 years after their diagnosis. Most cases of lung cancer would be prevented if people did not smoke tobacco products. Unfortunately, data on worldwide tobacco consumption suggest that lung cancer will remain an epidemic for years to come. Advances in understanding the pathogenesis, early detection, and therapy of lung cancer are in progress.

PREVALENCE

In 2007, approximately 213,380 new cases of lung cancer were diagnosed in the United States. Lung cancer is the second most frequently diagnosed cancer in both men and women; prostate and breast cancers are the most frequent in men and women, respectively (Fig. 1). The incidence of lung cancer peaked in men in 1984 (86.5/100,000 men) and has subsequently been declining (69.1/100,000 men in 1997). In women, the incidence increased during the 1990s, with a leveling off toward the end of the decade (43.1/100,000 women). These trends parallel the smoking patterns of these two groups.

Figure 1 Age-adjusted cancer incidence rates for male and female population, United States, 1975-2003.

Rates are per 100,000 population, adjusted to the 2000 U.S. standard population. The incidence of lung cancer in men and women has paralleled smoking habits.

(From Jemal A, Siegel R, Ward E, et al: Cancer statistics, 2007. CA Cancer J Clin 2007;57:43-66.)

Lung cancer is the leading cause of cancer-related mortality in both men and women. It surpassed colon cancer in the early 1950s in men and breast cancer in the late 1980s in women. Mortality rates in men declined significantly in the 1990s, whereas a slow increase occurred in women. These rates again parallel the smoking patterns of these two groups (Figs. 2 and 3). There were an estimated 160,390 deaths in 2007 in the United States secondary to lung cancer. This means that lung cancer accounts for approximately 29% of all cancer deaths. In men, lung cancer becomes the leading cause of cancer-related mortality from age 40 onward. In women, lung cancer surpasses breast cancer in those 60 years and older.¹

Figure 2 Age-adjusted cancer death rates for female population, United States, 1930-2003.

Rates are per 100,000 female population, adjusted to the 2000 U.S. standard population. Lung cancer surpassed breast cancer as the leading cause of cancer-related mortality in women in the late 1980s.

(From Jemal A, Siegel R, Ward E, et al: Cancer statistics, 2007. CA Cancer J Clin 2007;57:43-66.)

Figure 3 Age-adjusted cancer death rates for male population, United States, 1930-2003.

Rates are per 100,000 male population, adjusted to the 2000 U.S. standard population. Lung cancer has been the leading cause of cancer-related mortality in men since the early 1950s. Mortality rates began to decline in the 1990s.

(From Jemal A, Siegel R, Ward E, et al: Cancer statistics, 2007. CA Cancer J Clin 2007;57:43-66.)

RISK FACTORS

About 85% to 90% of patients with lung cancer have had direct exposure to tobacco. Many tobacco-related carcinogens have been identified; the two major classes are the N-nitrosamines and polycyclic aromatic hydrocarbons. A dose-response relation exists between the degree of exposure to cigarette smoke and the development of lung cancer. The age at which smoking began, the number of cigarettes smoked per day, and the duration of smoking all influence the likelihood of developing lung cancer. Also, the intensity of smoking, the depth of inhalation, and the composition of the cigarette influence the risk.

All cell types of lung cancer are associated with smoking. The strongest associations are with small cell and squamous cell carcinomas. The risk of developing lung cancer decreases over time after smoking cessation, although it never reaches that of a lifelong nonsmoker. Cigar smoking is also an independent risk factor for developing lung cancer.²

Exposure to sidestream smoke, or passive smoking, might lead to an increased risk of lung cancer. The risk varies with the level and duration of exposure. It is generally a much lower risk than is active smoking.³ Some suggest the risk is negligible.⁴

Many other risk factors have been identified (Box 1). Occupational agents are known to act as lung cancer carcinogens. Arsenic, asbestos, and chromium have the highest risk. An estimated 2% to 9% of lung cancers are related to occupational exposures. An inherited genetic predisposition has epidemiologic support as a risk factor, but the mechanisms are theoretical at this time.⁵ Women appear to have a higher baseline risk of developing lung cancer as well as a greater susceptibility to the effects of smoking. Differences in the metabolism of tobacco-related carcinogens and their metabolites or an effect of hormone differences are believed to account for the increased susceptibility.⁶

Dietary factors can modify risks. Higher consumption of fruits and vegetables is associated with a reduced lung cancer risk, and an increased dietary fat intake might lead to a higher risk. Supplementation with vitamins A and E, and beta carotene has not positively influenced risk.⁷

Chronic obstructive pulmonary disease is an independent risk factor. This risk increases as the forced expiratory volume in 1 second (FEV₁) decreases.^8,9

CLASSIFICATION

Pathologic features, visible on light microscopy, are used to categorize lung cancers. Lung cancers are divided into two major groups, small cell and non–small cell. These groups guide current evaluation and therapeutic decisions. The non–small cell cancer category consists of adenocarcinoma, squamous cell carcinoma, large cell carcinoma, and variants (Fig. 4).

Figure 4 Lung cancer histology.

Top left, Adenocarcinoma characterized by heterogeneous differentiation in the same tumor. Top right, Squamous cell carcinoma characterized by the presence of cytokeratin differentiation with keratinization and intercellular bridges. Bottom left, Large cell carcinoma characterized by sheets and nests with extensive necrosis, large nuclei with prominent nucleoli, and lack of definitive evidence of squamous or glandular differentiation. Bottom right, Small cell carcinoma characterized by round to fusiform nuclei, nuclear molding, faint or absent nucleoli, and scant cytoplasm.

Since the 1980s, the proportions of lung cancers that are adenocarcinomas and squamous cell carcinomas have changed. In North America, approximately 40% of all lung cancers are adenocarcinomas, and 20% to 25% are squamous cell. These figures were reversed in the past. The increased incidence of lung cancer in women (who are more likely to have adenocarcinomas) and changes in smoking habits are believed to account for this change.

PATHOPHYSIOLOGY

The pathophysiology of lung cancer development is complex and incompletely understood. The genes influenced in the pathogenesis of lung cancer produce proteins involved in cell growth and differentiation, cell cycle processes, apoptosis, angiogenesis, tumor progression, and immune regulation. Unveiling these mechanisms should translate into novel means of risk stratification, prevention, early detection, and therapy.

SIGNS AND SYMPTOMS

The clinical manifestations of lung cancer result from the effects of local growth of the tumor, regional growth or spread through the lymphatic system, hematogenous distant metastatic spread, and remote paraneoplastic effects from tumor products or immune cross-reaction with tumor antigens (Box 2).

Local growth in a central location can cause cough, hemoptysis, or features of large-airway obstruction. Peripheral growth can also cause cough and dyspnea. If the pleura or chest wall becomes involved, pain can occur. Regional growth can lead to esophageal compression (dysphagia), recurrent laryngeal nerve paralysis (hoarseness), phrenic nerve paralysis with an elevated hemidiaphragm (dyspnea), and sympathetic nerve paralysis leading to Horner’s syndrome (ptosis, miosis, anhidrosis, and enophthalmos). Apical growth can lead to Pancoast’s syndrome, with shoulder pain radiating in an ulnar distribution. The superior vena cava can become obstructed and the heart and pericardium can become involved. Lymphatic obstruction and spread can lead to dyspnea, hypoxia, and pleural effusions.

Distant metastatic disease can affect most organs. Neurologic symptoms can suggest brain metastases or spinal cord compression, and pain could indicate bone metastases. Laboratory abnormalities can point to bone marrow or liver involvement. Imaging might detect adrenal involvement.

Paraneoplastic syndromes can occur before the primary tumor appears and thus can be the first sign of disease or an indication of tumor recurrence. Paraneoplastic endocrine syndromes occur when the tumor produces hormones. The three most common are ectopic Cushing’s syndrome, the syndrome of inappropriate antidiuretic hormone (SIADH), and humoral hypercalcemia of malignancy. Ectopic Cushing’s syndrome occurs in 2% to 10% of patients with small cell carcinoma. The clinical manifestations are less prominent than in Cushing’s disease; biochemical abnormalities predominate, whereas the physical changes are less prominent. The SIADH is also more common in small cell carcinoma, occurring in 7% to 11% of patients. The manifestations of hyponatremia (mental status changes, lethargy, or seizures) are often absent despite very low sodium levels, because the rate of decline is typically prolonged. Humoral hypercalcemia of malignancy, resulting from the production of parathyroid hormone-related protein by the tumor, is most commonly associated with squamous cell carcinoma. Fatigue, mental status changes, weakness, gastrointestinal symptoms, polyuria, and electrocardiogram changes may occur.

Paraneoplastic neurologic syndromes affect all parts of the nervous system. An immune response to tumor antigens that cross-react with common antigens expressed in the nervous system seems to take place. This leads to manifestations that vary depending on where in the nervous system these antigens are expressed. Paraneoplastic limbic encephalitis is characterized by mood and behavior changes, memory problems, and seizures; paraneoplastic cerebellar degeneration manifests with ataxia, nystagmus, dysarthria, and diplopia; and paraneoplastic opsoclonus-myoclonus manifests with involuntary eye movements, myoclonus, truncal ataxia, dysarthria, and encephalopathy. Each of these is more common with small cell carcinoma, can occur in the presence of anti-Hu antibodies, and can occur as part of a more diffuse anti-Hu syndrome (the encephalomyelitis and subacute sensory neuropathy syndrome).

Other paraneoplastic neurologic syndromes include cancer-associated retinopathy and the Lambert-Eaton myasthenic syndrome. In cancer-associated retinopathy (most common with small cell carcinoma), rapid vision loss, ring scotomata, photosensitivity, night blindness, and color vision loss can occur in association with autoantibodies directed against retinal proteins. Lambert-Eaton myasthenic syndrome is the most common of the neurologic paraneoplastic syndromes and is present in 3% of small cell carcinomas. Proximal muscle weakness (which might improve with exercise) is most prominent in the lower extremities, and autonomic features predominate. Autoantibodies directed against P/Q type voltage-gated calcium channels are believed to be responsible.

Other paraneoplastic syndromes include skeletal and connective tissue syndromes (clubbing, hypertrophic pulmonary osteoarthropathy), coagulation and hematologic disorders, cutaneous and renal manifestations, and systemic symptoms (anorexia, cachexia, and weight loss).¹⁰

DIAGNOSIS

Approximately 85% of patients with lung cancer are symptomatic at presentation. In the remainder, lung cancer is detected by radiographic evaluation initiated for an unrelated problem. This proportion might change in the future if currently investigated screening techniques prove beneficial. Chest radiography and computed tomography (CT) are performed at most patients’ initial evaluation. Clinical and radiographic features of the presentation dictate further evaluation.

Clinical features that suggest malignancy on initial evaluation include older age, current or past history of tobacco abuse, hemoptysis, and the presence of a previous malignancy. Radiographic features suggesting malignancy include the absence of a benign pattern of calcification in the detected lesion, a nodule or mass that is growing, a nodule with a spiculated or lobulated border, a larger lesion (>3 cm is considered malignant unless proven otherwise), and a cavitary lesion that is thick walled. Modern imaging techniques are used to alter the clinical probability of malignancy and hence influence biopsy decisions. Positron emission tomography (PET) using ¹⁸F fluorodeoxyglucose is the most-studied ancillary imaging technique. It has a sensitivity of 97% and a specificity of 78% as used in clinical practice.¹¹ Single-photon emission CT and lung nodule enhancement with contrast-enhanced CT are less well established.

Ultimately, tissue needs to be obtained to confirm the diagnosis of lung cancer. Flexible bronchoscopy and transthoracic needle biopsy are the invasive, nonsurgical approaches used to obtain tissue. If they fail or are deemed unnecessary, a surgical approach is used.

Flexible bronchoscopy has a high diagnostic yield for endoscopically visible lesions. The addition of endobronchial needle aspiration to conventional sampling techniques (washing, brushing, and endobronchial biopsy) improves this yield. The diagnostic yield from peripheral lesions is lower. Conventional sampling techniques and peripheral transbronchial needle aspiration complement each other. Factors that influence the diagnostic yield of flexible bronchoscopy for peripheral lesions include the size of the lesion, its location, and a bronchus sign on CT. Smaller, more peripheral lesions, without a visible bronchus within or leading directly to them, are unlikely to be diagnosed by flexible bronchoscopy.

Transthoracic needle biopsy, using fluoroscopic or CT guidance, can be used to obtain tissue. The positive predictive value of this procedure is high, the negative predictive value is modest, and the rate of establishing a specific benign diagnosis is low. Smaller nodules in central locations have lower diagnostic rates. A higher rate of pneumothorax occurs with transthoracic needle biopsy; thus, flexible bronchoscopy is often attempted first.¹²